UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                                WASHINGTON, D.C. 20460
        mo**
                                                                OFFICE OF THE ADMINISTRATOR
                                                                  SCIENCE ADVISORY BOARD

                                   January 25, 2006

EPA-SAB-06-002

The Honorable Stephen L. Johnson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460

       Subject: Review of EPA's Draft Framework for Inorganic Metals Risk Assessment

Dear Administrator Johnson:

   The Environmental Protection Agency's Office of Research and Development requested that
the Science Advisory Board (SAB) review the Agency's draft Framework for Inorganic Metals
Risk Assessment (the Framework). The Framework was developed to supplement previous EPA
guidance for risk assessment activities related to metals. A panel of the SAB reviewed the
Framework and has commented on the state of the science presented in the document, as well as
the recommendations, supporting tools, methods,  and models. The enclosed SAB report
addresses EPA's  charge questions to the Panel and provides recommendations to improve the
Framework.

   The SAB commends EPA for initiating the development of a comprehensive risk assessment
framework for metals and metalloids. The SAB finds that the Framework covers the main areas
of concern to risk assessors. However, the SAB also finds that technical corrections and
additions are needed and that the document should be restructured and substantially revised to
improve the clarity of expression, precision of wording, and balance and depth of coverage of
important topics. In this regard, the SAB finds that:

   *   The purpose of the Framework is unclear. The document attempts to serve as a
       description of basic scientific principles as well as a practical guide for risk assessors. To
       serve these two purposes, the document requires revision to provide a more balanced
       presentation of scientific principles and risk assessment guidance. The document should
       also clearly differentiate the following: the framework for assessment; examples to
       illustrate and clarify framework issues', and specific instructions for risk assessors.

-------
    •   The scientific synthesis in the Human Health and Ecological Sections of the Framework
       is incomplete and in need of major revision.  Important scientific issues in other parts of
       the Framework are either missing or lack clarity. Specific SAB comments and
       recommendations are provided to address these concerns.

    •   The Framework provides comprehensive coverage of available tools and methods for
       metals risk assessment. However, critical evaluations of tools and methods are
       sometimes unbalanced or lacking. The Framework should focus on the strengths,
       weaknesses, and limitations of various methods and tools. Where appropriate,
       comparative assessment of competing approaches should be provided.

    •   The Recommendations Section of the Framework should be revised to reduce the overall
       number of recommendations by combining redundancies and eliminating those
       statements that are not recommendations. Recommendations in the Framework should
       also be organized according to their specificity (i.e., from general overarching to more
       specific), and each recommendation should be adequately supported by text and
       references as appropriate.

   In summary, the SAB finds that the Framework for Inorganic Metals Risk Assessment is an
important document that will guide EPA and others in evaluating meta!s in ecological and human
health risk assessment.  Revision of the Framework is necessary before it is published in final
form in order to make it of more current and long-term value to EPA. The SAB strongly urges
EPA to continue developing the Framework and has provided specific comments and
recommendations to improve the document. The SAB is willing to provide additional review of
the revision of the Framework.

                                     Sincerely,
      Dr. M. Granger Morgan. Chair
      EPA Science Advisory Board
            5^t__
Dr. Deborah L. Swackhamer, ChaiT"~~"~~~
Metals Risk Assessment
Framework Review Panel
EPA Science Advisory Board

-------
r
                                                      NOTICE

                     This report has been written as part of the activities of the EPA Science Advisory Board,
                  a public advisory group providing extramural scientific information and advice to the
                  Administrator and other officials of the Environmental Protection Agency. The Board is
                  structured to provide balanced, expert assessment of scientific matters related to the problems
                  facing the Agency.  This report has not been reviewed for approval by the Agency and,
                  hence, the contents of this report do not necessarily represent the views and policies of the
                  Environmental Protection Agency, nor of other agencies in the Executive Branch of the
                  Federal government, nor does mention of trade names or commercial products constitute a
                  recommendation for use. Reports of the EPA Science Advisory Board are posted on the EPA
                  website at http://www.epa.gov/sab.
                                                         in

-------
                       U.S. Environmental Protection Agency
                               Science Advisory Board
               Metals Risk Assessment Framework Review Panel
 CHAIR
 Dr. Deborah L. Swackhamer, Professor, Division of Environmental Health Sciences, School of
 Public Health, and Co-Director, Water Resources Center, University of Minnesota, Minneapolis,
 MN

 MEMBERS
 Dr. Max Costa, Professor and Chairman, Department of Environmental Medicine, New York
 University School of Medicine, New York, NY

 Dr. David Dzombak, Professor, Department of Civil and Environmental Engineering, Carnegie
 Mellon University, Pittsburgh, PA

 Dr. Kevin Farley, Professor, Department of Civil and Environmental Engineering, Manhattan
 College, Riverdale, NY

 Dr. Ivan Fernandez, Professor, Department of Plant, Soil, and Environmental Sciences,
 University of Maine, Orono, ME

 Dr. Bruce Fowler, Assistant Director for Science, Division of Toxicology, Agency for Toxic
 Substances and Disease Registry, Atlanta, GA

 Dr. Andrew J. Friedland, Professor and Chair, Environmental Studies Program, Dartmouth
 College, Hanover, NH

 Dr. A. Jay Gandolfi, Assistant Dean for Research and Graduate Studies, College of Pharmacy,
 University of Arizona, Tucson, AZ

 Dr. Joshua Hamilton, Professor, Department of Pharmacology and Toxicology, Dartmouth
 Medical School, Hanover, NH

 Dr. Kim Hayes, Professor and Director, Environmental and Water Resources Engineering
 Program, University of Michigan, Ann Arbor, MI

 Dr. Robert Hudson, Associate Professor, Department of Natural Resources and Environmental
 Science, University of Illinois at Urbana-Champaign, Urbana, 1L

Dr. Thomas La Point, Professor and Director, Department of Biological Sciences, University of
North Texas, Denton, TX

Dr. Samuel Luoma, Senior Research Hydrologist, U.S. Geological Survey, Menlo Park, CA

-------
r
               Dr. Glenn Miller, Director, Center for Environmental Science and Engineering, University of
               Nevada, Reno, NV

               Dr. James Shine, Assistant Professor of Aquatic Chemistry, Department of Environmental
               Health, School of Public Health, Harvard University, Boston, MA

               Dr. Katherine Squibb, Associate Professor, Department of Epidemiology and Preventative
               Medicine, University of Maryland School of Medicine, Baltimore, MD

               Dr. William Stubbleiield, Senior Environmental lexicologist, Parametrix, Inc., Albany, OR

               Dr. Bernard Weiss, Professor of Environmental Medicine, University of Rochester Medical
               Center, Rochester, NY

               Dr. John Westall, Professor, Department of Chemistry, Oregon State University, Corvalis, OR

               Dr. Herbert Windom, Professor, Skidaway Institute of Oceanography, Savannah, GA

               Dr. Judith Zeiikoff, Associate Professor, Department of Environmental Medicine, New York
               University School of Medicine, Tuxedo, NY

               SCIENCE ADVISORY BOARD STAFF
               Dr. Thomas Armitage, Designated Federal Officer, U.S. Environmental Protection Agency,
               Washington, DC

-------
                                 TABLE OF CONTENTS


 1.   EXECUTIVE SUMMARY	viii


 2.   INTRODUCTION	1


 3.   CHARGE TO THE REVIEW PANEL	2


 4.   REVIEW PROCESS	4


 5.   OVERARCHING COMMENTS AND RECOMMENDATIONS	5


 6.   RESPONSE TO THE CHARGE QUESTIONS	12

 6.1.1     Charge Question 1.1	12
  6.1.1.1   Comments in Response to Charge Question 1.1	12
  6.1.1.2   Key Recommendations in Response to Charge Question 1.1	13

 6.1.2    Charge Question 1.2	13
  6.1.2.1   Comments in Response to Charge Question 1.2	14
  6.1.2.2   Key Recommendations in Response to Charge Question 1.2	14

 6.2.1     Charge Question 2.1	15
  6.2.1.1   Comments in Response to Charge Question 2.1	15
  6.2.1.2   Key Recommendations in Response to Charge Question 2.1	18

 6.2.2     Charge question 2.2	19
  6.2.2.1   Comments in Response to Charge Question 2.2	19
  6.2.2.2   Key Recommendations in Response to Charge Question 2.2	21

 6.3.1    Charge Question 3.1	21
  6.3.1.1   Comments in Response to Charge Question 3.1	21
  6.3.1.2   Key Recommendations in Response to Charge Question 3.1	23

 6.3.2    Charge Question 3.2	24
  6.3.2.1   Comments in Response to Charge Question 3.2	24
  6.3.2.2   Key Recommendations in Response to Charge Question 3.2	27

 6.3.3    Charge Question  3.3	29
  6.3.3.1   Comments in Response to Charge Question 3.3	29
  6.3.3.2   Key Recommendations in Response to Charge Question 3.3	31

6.3.4     Charge Question 3.4	32
  6.3.4.1   Comments in Response to Charge Question 3.4	32
  6.3.4.2   Key Recommendations in Response to Charge Question 3.4	34

6.3.5     Charge Question 3.5	35
  6.3.5.1   Comments in  Response to Charge Question 3.5	35
  6.3.5.2   Key Recommendations in Response to Charge Question 3.5	36

6.3.6      Charge Question 3.6	36
  6.3.6.1   Comments in  Response to Charge Question 3.6	36

-------
  6.3,6.2   Key Recommendations in Response to Charge Question 3.6	37

6.3.7     Charge Question 3.7	38
  6.3.7.1   Comments in Response to Charge Question 3.7	38
  6.3.7.2   Key Recommendations in Response to Charge Question 3.7	39

6.3.8    Charge Question 3.8	39
  6.3.8.1  Comments in Response to Charge Question 3.8	39
  6.3.8.2  Key Recommendations in Response to Charge Question 3.8	41

6.3.9     Charge Question 3.9	42
  6.3.9.1  Comments in Response to Charge Question 3.9	42
  6.3.9.2  Key Recommendations in Response to Charge Question 3.9	43

6.3.10    Charge Question 3.10	43
  6.3.10.1 Comments in Response to Charge Question 3.10	43
  6.3.10.2 Key Recommendations in Response to Charge Question 3.10	44

6.3.11    Charge Question 3.11	45
  6.3.11.1 Comments in Response to Charge Question 3.11	45
  6.3.11.2 Key Recommendations in Response to Charge Question 3.11	49

6.3.12    Charge Question 3.12	50
  6.3.12.1 Comments in Response to Charge Question 3.12	51
  6.3.12.2 Key Recommendations in Response to Charge Question 3.12	51

6.3.13    Charge Question 3.13	52
  6.3.13.1 Comments in Response to Charge Question 3.13	52
  6.3.13.2 Key Recommendations in Response to Charge Question 3,13	53

6.3.14    Charge Question 3.14	53
  6.3.14.1 Comments in Response to Charge Question 3.14	53
  6.3.14.2 Key Recommendations in Response to Charge Question 3.14	54


7.  REFERENCES	56


APPENDIX A. DETAILED COMMENTS AND SUGGESTED TECHNICAL CORRECTIONS IN RESPONSE
TO CHARGE QUESTION 3.1	A-l


APPENDIX B. SPECIATION	B-l


APPENDIX C. SUGGESTED EDITORIAL/WORDING CHANGES	C-l
                                                VII

-------
 I.
EXECUTIVE SUM1S1ARY
   The U.S. Environmental Protection Agency (EPA) Science Advisory Board (SAB) Metals
Risk Assessment Framework Review Panel has reviewed EPA's draft Framework for Inorganic
Metals Risk Assessment (the Framework).  This report transmits the SAB's comments and
recommendations. Many EPA programs face decisions on whether and how to regulate metals.
These decisions range from setting standards or permits for environmental releases to
establishing safe levels in different environmental media, to setting priorities for programmatic
or voluntary efforts. EPA developed the draft Framework for Inorganic Metals Risk Assessment
to supplement previous Agency guidance for use in site-specific risk assessments, criteria
derivation, and other similar Agency activities related to metals.

   EPA followed a stepwise process to develop the Framework document. A Metals Action Plan
was first developed to provide brief descriptions of EPA's current activities related to metals and
to identify critical scientific issues to be addressed. After the Metals Action Plan was reviewed
by the Science Advisory Board, issue papers were developed to discuss key scientific topics
pertaining to inorganic metals. The draft document, Framework for Inorganic Metals Risk
Assessment was then completed and submitted to the SAB for review. The Framework
document contains five sections: i) an introduction describing the purpose, audience, and scope
of the document; 2) a problem formulation section setting forth principles of metals risk
assessment and providing a conceptual model for metals assessments; 3) a section providing
recommendations  that should be considered when conducting assessments of the ecological or
human health risks associated with metals exposures; 4) a section on metal specific topics and
methods in environmental chemistry, human health exposure pathway analysis, human health
effects, ecological exposure pathway analysis, and characterization of ecological effects; and 5) a
section identifying metals research needs.

   EPA sought comment from the SAB on the scientific soundness of the Framework's synthesis
and representation of the state of the science.  Specifically, EPA sought comment on the overall
objectivity and utility of the recommendations and supporting tools, methods, and models to its
primary audiences (EPA risk assessors and the public), and whether there were any additional
research needs that warrant inc usion or further discussion in the Framework. EPA defined
objectivity as: "a focus on whether the disseminated information is being presented in an
accurate, clear, complete, and unbiased manner, and as a matter of substance, is accurate,
reliable, and unbiased." EPA defined utility as: "the usefulness of the information to its intended
users, including the public."

   The SAB notes that the Framework will be an important document.  It will be used by EPA to
develop more detailed risk assessment guidance, and  it will be used by both EPA and the
external community as an authoritative compilation of the state of science regarding metals in the
environment The SAB commends EPA for initiating the development  of a risk assessment
framework for metals that covers a broad spectrum of topics related to human health and
ecological risk concerns associated with exposure to toxic metals and metalloids. The SAB finds
that the Framework clearly identifies the unique attributes of managing  metals.  However, the
SAB also finds that a number of major issues within the Framework document need to be
addressed. In order to make the Framework of long-term value to EPA, significant revision of
                                          VIII

-------
the document is required before it is published in final form. The SAB notes that the overall
clarity of expression, precision of wording, and balance in coverage among topics in the
Framework must be greatly improved. In response to EPA's charge questions, the SAB provides
specific comments and recommendations for improvements in the Framework.  In addition, the
SAB has noted other recommendations for  improvement that are overarching in scope and not
related to specific charge questions. Because of the scope of revisions recommended, the SAB
believes the revised Framework would benefit from a second external peer review. The SAB is
willing to provide such a review.

Overall Framework Scope (Charge question 1.1)

   •   The SAB generally finds that the overall Framework scope is sufficiently broad  and
       provides an appropriate level of flexibility in addressing issues of concern. The  SAB
       supports the idea of treating both human health and ecological risks in one document in
       order to consistently present risk assessment concepts.  However, a major weakness in
       the current version of the Framework is the lack of consistency in identity. The
       Framework appears to vacillate between being a description of basic principles to a
       methods manual. The SAB therefore recommends that the Framework be reviewed and
       revised to remove any confusion in  its intended purpose.  If the document is to serve as
       both a framework and a practical guide for risk assessors, the recommendations  and
       guidance in the document should be balanced and organized consistently with this dual
       purpose in mind. The SAB recommends that EPA clearly identify and carefully
       differentiate material that is presented as "the framework for assessment," "examples to
       illustrate and clarify framework issues," and "specific instructions,"

General Risk Assessment Categories in the Framework (Charge question 1.2)

   •   The SAB generally finds that the risk assessment categories listed in the introduction of
       the Framework are an appropriate context to cast the relevant issues of metals in
       comparison to organic compounds.  However the SAB recommends that the scope of the
       assessment categories be more clearly defined, and that the number of assessment
       categories be expanded to span the range of complexity among screening and site-
       specific risk assessments conducted at different scales. The SAB also finds that  the
       sections of the Framework following the introduction largely concern site-specific
       assessment  issues.  The SAB recommends that the subsequent sections of the document
       be revised to represent more balance among the different types of assessments.

Articulation/Objectivity of Metals Assessment Principles (Charge question 2.1)

   •   The SAB notes that the framework discusses factors to be considered in  metals risk
       assessment  rather than principles. The SAB therefore recommends that EPA use the
       words "factors to be considered" or "factors" in Section 2 of the Framework instead of
       "principles." The SAB finds that there is an imbalance in coverage of factors in the
       Framework and recommends inclusion in the document of a number of important factors
       such as nature and type of metals source, route of metals exposure, and involvement of
       metals in biogeochemical cycles. The SAB also recommends that EPA list key  questions
                                          IX

-------
       for all of the factors discussed in the Framework. The discussion associated with the key
       questions should identify why the factors are uniquely important for metals risk
       assessment.

Conceptual Model (Charge question 2.2)

    •   The SAB finds that the; conceptual model in the Framework is sufficiently
       comprehensive. However the SAB recommends that the model be revised to emphasize a
       number of key concepts discussed in the response to charge question 2.2 below, and to
       more clearly distinguish differences between metal/metalloid and organic pollutants. The
       SAB also recommends that the conceptual model be more clearly linked to the related
       discussion in various parts of the Framework.

Recommendations in the Framework (Charge question 3.1)

    •   The SAB has identified revisions needed to address technical issues concerning the
       recommendations section of the Framework (Section 3). Specific revisions are suggested
       in the response to charge question 3.1 below. The SAB finds that the clarity of the
       framework could be improved by organizing the recommendations according to their
       specificity (i.e., from the most critical general overarching recommendations with the
       greatest impact to more specific recommendations of interest to the assessor).  The SAB
       recommends that the Framework be revised to reduce the number of recommendations in
       the document by combining those that are redundant or similar. It is also recommended
       that prescriptive recommendations be generalized or cited as examples of appropriate
       applications of metals principles.  EPA should also review the Framework and make
       necessary revisions to ensure that the recommendations are expressed as
       recommendations, not simply factual statements.

Objectivity and Utility of the Data, Tools, and Methods in Section 4 of the Framework (Charge
question 3.2)

    •   The SAB finds that the human exposure and health effects discussion in Section 4 of the
       Framework is incomplete and in need of major revision. For example, the Framework is
       incomplete in the following areas:

          1.  Treatment of particulate matter. Particulate matter less than 2.5 micrometers in
             size and nanoparticles are of critical concern for the exposure and delivery of
             metals to humans and this is ignored in the Framework.

          2.  Characteristics of inhaled particles. The characteristics of inhaled particles are
             critical determinants of metals uptake in humans and this is not addressed in the
             Framework.
          3.   Inhaled mixtures. Biological effects of inhaled mixtures such as metals in
             combination with other airborne pollutants are not addressed in the Framework.

-------
         4. Dermal exposure and uptake of metals.  Dermal exposure and uptake of metals are
            not adequately addressed in the Framework.

         5. Effects of metals at low doses. The Framework does not provide a discussion of
            the toxic effects of metals at low doses.

         6. Interactions between metals and organic chemicals.  There is insufficient
            discussion in the Framework of the interactions between metals and organic
            chemicals and how these interactions can lead to potentiation or antagonism.

        The SAB recognizes that a rewrite addressing areas of incompleteness may not be
        achievable in the short-term, but it will be essential if the treatment of human exposure
        and health effects is to be of equal value and  quality to other parts of the Framework.
        Recommendations to improve the human exposure and health effects discussion are
        provided in the responses to charge question  3.2 and in the detailed comments  in
        Appendices A and C.

   •  The SAB finds that the environmental chemistry discussion in Section 4 of the
      Framework is comprehensive, but in many instances critical evaluations of the tools and
      methods are not provided, and the justification  for many recommendations is not clear.
      As discussed below, the SAB recommends that more emphasis be placed on developing
      comparative assessments of available tools and methods.

   •  The SAB finds that the ecological exposure and effects discussion in Section 4 of the
      Framework provides a great deal of supporting information for the recommendations
      articulated in the document.  However, the treatment of various topics addressed  in the
      ecological exposure and effects section is uneven and leaves the impression of not being
      objective. In some places, the discussion does  not fully reflect the state of the science.
      The SAB recommends that the bioaccumulation and bioavailability sections of the
      Framework treat the routes of exposure (diet and dissolved  metals) in an integrated
      fashion. This could be accomplished by organizing the discussion around the
      bioavailability conceptual model. The SAB recommends that the toxicity testing section
      of the Framework discuss uncertainties, such as the lack of dietary exposure in laboratory
      toxicity tests, that are of particular importance to metals risk assessment. The SAB
      recommends that the discussions of sediment contamination be revised to address
      important principles and methods that are currently absent.  The SAB also recommends
      that the discussion of simultaneously extracted metals-acid  volatile sulfides (SEM-AVS)
      be revised to capture the controversies surrounding this approach.  In addition, the
      discussion of the biotic ligand model (BLM) should address the limits of the approach
      and its early state of development.

Metals Speciation (Charge question 3.3)

   •  The SAB commends EPA for emphasizing the  concept of metals speciation in the
      Framework. 1 lowever, the SAB finds that a clear definition of the terms species  and
      speciation should be included in Section 2 of the document. The SAB provides such

-------
       definitions in Appendix B of this report.  As discussed in the response to charge question
       3.3 below, the SAB also finds that the treatment of speciation in the Framework could be
       improved by providing more accurate and detailed information. The SAB finds that the
       value of some approaches to considering speciation is overstated in the Framework (e.g.,
       application of the biotic ligand model to chronic or natural exposures). Other approaches
       to considering speciation are ignored in the Framework (e.g., direct measurement of
       speciation). The SAB recommends that appropriate linkages between speciation and the
       concepts used in risk analysis, such as partitioning and bioavailability, be emphasized in
       the Framework. In addition, it would be helpful to include a fuller description of the
       currently available tools to quantify metal speciation in environmental samples, including
       the strengths and weaknesses of each technique.  The SAB notes that the paucity of data
       to support modeling of speciation limits the risk assessor's ability to include speciation in
       metal risk assessment tasks at site and national scales. However, because of
       transformations that occur as metals cycle through the environment, metal speciation
       determination is more applicable for site-specific investigations than the setting of
       national standards.

Summary Recommendations Tables in the Framework (Charge question 3.4)

   •   The SAB finds that summary recommendation tables such as example Table A-l in the
       Framework can be used to effectively present important recommendations in an
       organized manner. As discussed in the response to charge question 3.4, the SAB
       recommends that the tables be restructured to relate the recommendations to the
       categories of risk assessment discussed in the document.

Environmental Chemistry (Charge questions 3.5 — 3.7)

   •   Objectivity of Hard Soft Acid Base Concept. The SAB finds that the application of the
       Hard Soft Acid Base concept to the stability of metal complexes in the general context of
       risk assessment is presented in an unbiased manner.  However, the SAB recommends that
       general statements in the Framework indicating that hard acids are more toxic than soft
       acids should be worded more carefully.  The SAB notes that the Hard Soft Acid Base
       concept is useful for assessing the relative strength of binding of a metal to a receptor, but
       the toxic response to bound metal is not adequately addressed by the Hard Soft Acid Base
       concept.

   •   Objectivity of Atmospheric Metal Chemistry Discussion.  The SAB notes that none of the
       Metals Risk Assessment Framework Review Panel members has an active research
       program in atmospheric chemistry.  The SAB therefore recommends that an atmospheric
       chemist review the atmospheric chemistry sections of the Framework to ensure that there
       are no gaps in coverage beyond those identified in the response to charge question 3.6.

   •   Objectivity of the Chemistry and Environmental Parameters in Metal Surface
       Complexalion and Partition Coefficient Models. The SAB finds the Framework
       discussion of surface ccmplexation models to be generally accurate and unbiased.
       However, in response to charge question  3.7 below the SAB identified a number of areas
                                          XII

-------
       where the presentation lacks completeness. The SAB has provided recommendations to
       improve these sections of the document.

Human Exposures and Effects (Charge questions 3.8-3.10)

   •   Objectivity of the Discussion on Natural Background of Metals.  The SAB recommends
       that EPA use the term "ambient" or "ambient levels" in the Framework rather than
       "background." The SAB also recommends defining and using the terms "body burden"
       and "human biological monitoring" in the Framework glossary and text. The SAB finds
       that the term "background" is often incorrectly assumed to connote natural and therefore
       safe, or of no significant human or ecological health concern. However, ambient levels
       can vary, or can be inherently high enough to represent a potential health concern by
       themselves. Ambient levels can also represent a total concentration from a combination
       of natural and anthropogenic sources, some of which may be historical or unknown. The
       SAB acknowledges, however, that the term "background" exposure has been used in
       human health dose response assessment.  This term, referring to both the natural and
       anthropogenic sources of a chemical under assessment, has been used in various
       publications on dose-response modeling over the past 25 years or more.

   *   Objectivity of the Discussion of Essentiality Versus Toxicity. The SAB finds that
       revisions are needed in the Framework to clarify and ensure accuracy of the discussion of
       essentiality  versus toxicity. The SAB recommends that EPA carefully define
       "essentiality," recognize that metals essential to some organisms may not be essential to
       others, recognize that essential metals can cause adverse health effects at elevated
       concentrations, and recognize that the source and route of exposure play an  important
       role in the toxicity of essential metals.

   •   Objectivity of the Discussion and Recommendations for Assessing the Toxicity of
       Mixtures. The SAB finds that the Framework requires revision to more explicitly address
       a number of issues concerning metals mixtures.  The SAB recommends that the
       Framework be revised to address: competitive interactions among chemically similar
       metals/metalloids (mimicry), reduction of metal  reactivity and increase in mobility by
       organic compounds that form complexes with metals, and possible increases in toxic
       effects for organic compounds that form  lipophilic complexes with metals.

Ecological Exposures and Effects (Charge questions 3.11-3.14)

   •   Objectivity  of the Discussion and Recommendations Concerning Natural Background,
       Bioavailabiiity,  Bioaccumulation, Biomagnification, and Trophic Transfer.  The SAB
       finds that the Framework discussions of natural background, bioavailability,
       bioaccumulation, biomagnification, and trophic transfer require revision to address
       inconsistencies between Sections 3 and 4 of the document. In the response to the charge
       question 3.11, the SAB recommends specific revisions to integrate the sections, address
       imbalance among the recommendations,  integrate discussions of uncertainties, and
       address omissions.
                                          Xlll

-------
    •  Objectivity of the Framework Discussion Concerning the use of Bioconcentration factor
       (BCF) and Bioaccumulation Factor (BAF). The SAB agrees with the statement in the
       Framework indicating that BCF/BAF methodologies are not good measures of hazard for
       metals. However, the SAB finds that a clearer and more; systematic discussion is needed
       in the document to jus :ify this statement. The SAB recommends that EPA revise the
       Framework to include a discussion of what could replace BCF/BAF as a measure of
       bioaccumulative potential and where BCF/BAF approaches are useful.

    •  Derivation of Bioaccumulation Factors (BAFs) and Bioconcentration Factors (BCFs).
       The SAB finds that the mathematical relationships in the Framework appropriately
       represent the metals concentration in the organism or tissue as a function of the
       bioavailable concentration in the exposure medium/media for each set of exposure
       conditions. However, the SAB recommends that in the future. EPA incorporate a
       bioenergetics approach into the Framework.  Such an approach offers valuable potential
       for understanding metal accumulation from air, sediments, soils, or water.  In the interim,
       the SAB recommends that the Framework address metals bioaccumulation empirically
       for site assessments.

    •  Objectivity of Information and Recommendations Concerning use of Simultaneously
       Extracted Metal-Acid Volatile Sulfides (SEM-AVS) Approach and the Biotic Ligand
       Model (ELM).  The SAB finds that the Framework comprehensively describes the theory
       and evidence supporting the use of the SEM-AVS approach and the BLM. However, as
       further discussed in the response to charge question 3.14 below, the SAB finds that the
       Framework is unbalanced in presenting the practical and theoretical challenges and
       inherent limitations encountered in the use of these methods. The SAB recommends that
       the Framework be revised to provide a more balanced presentation of the "pros and cons"
       associated with the metnods.

Additional Major Revisions

    •  Title.  The SAB finds that the title  of the Framework is awkward: metals are inorganic by
       definition, and thus the use of the adjective "inorganic" in front of metals is redundant.
       Although the SAB realizes that the adjective "inorganic" was probably used in the
       original title to exclude organometallics, especially methylmercury, it detracts from the
       clarity of the title.  A better title would be "Framework for Assessment of Risk of Metals
       and Metalloids in the Environment." At the beginning of the Framework document, EPA
       should clarify the rationale for specifically excluding organometallic compounds from the
       document while including natural transformation pathways that form organometallic
       compounds.

    •   Balance and Comparability Among Sections.  The SAB finds that Section 3 of the
       Framework should be reorganized  and rewritten to provide more comparability among
       the discussions of human health effects, aquatic effects, and terrestrial effects. The  SAB
       recommends that the aquatic section be modeled on the terrestrial section.
       Recommendations should be highlighted by cross-referencing justification to Section 4 of
       the Framework and minimizing textual justification.

-------
Restructuring of Framework Document. The SAB recommends that Section 4 of the
Framework be reorganized to mirror the organizational structure used in Section 3.

Illustrative Examples.  The SAB finds that illustrative examples would be useful
throughout the document. Examples of how certain recommendations might be
implemented would greatly improve the utility of the document.

Discussion of Uncertainties and Data Quality. The SAB finds that the discussion of
uncertainties of tools, methods and data is generally lacking and inconsistent throughout
the Framework. The SAB recommends that the importance of critically considering data
quality be explicitly stated throughout the document wherever the use of analytical data is
discussed.

Use of the term "Bioaccumulation " versus  "Accumulation " to Describe Metals
Concentrations. It is the opinion of the SAB that there should not be a distinction in the
Framework between the term "bioaccumulation" to describe metal concentration in
aquatic and terrestrial organisms and the term "accumulation" of metals for humans.
This is not an accepted distinction in the scientific community.  In humans as in other
terrestrial animals, the steady-state body burden of many metals is under homeostatic
control that balances intake and excretion. However, for certain metal compounds
bioaccumulation can occur, which can be defined as either a persistent increase in
individual steady-state levels that is correlated with higher prior exposure, and/or a
progressive increase in body burden as a function of exposure time or age, that  is above
normal steady-state levels and which may involve selective bioaccumulation of the metal
in certain tissues. The SAB believes it is important to recognize that some metals do
bioaccumulate in the tissues of humans and that this bioaccumulation is related to their
toxicity.  To clarify what is meant by bioaccumulation, the SAB recommends that the
definition of the term "bioaccumulation" in the glossary of this  document be modified to
read as follows:

Bioaccumulation:  The net accumulation of a metal in a tissue of interest or the whole
organism that results from exposure to all environmental sources,  including air, water,
solid phases (i.e., soil, sediment) and diet, and that represents a net mass balance
between  uptake and elimination of the metal.

Metal-specific Reference  Values (R/D/RfC) and/or Cancer Potency Factors. The SAB
recommends that, in introducing the Human Health Effects Section, EPA set the context
by explaining that human health risk assessors start their analysis with a metal-specific
reference value (RfD/RfC) and/or cancer potency factor that has been developed through
a process separate from the risk assessment. The role of the human risk assessor is to
appropriately integrate the reference values and potency factors with the exposure
assessment.  Thus, the risk assessor needs an understanding of the toxicological endpoints
and mechanisms of action that underlie the derivation of these values to ensure that, for
example, the appropriate population and life stages are addressed, appropriate dietary
aspects are taken into consideration, and the appropriate exposure pathways are
                                    XV

-------
       considered. For metals, frequency and duration of exposure, as well as exposure
       concentrations, are important parameters to be considered for accurate dose assessments.
       The Framework should focus on advising human health risk assessors on how to take
       these considerations into account in constructing the risk assessment.

    •  Modeling, The SAB notes that the Framework accurately reflects the fact that modeling
       the environmental fate and transport of metals differs in significant ways from modeling
       organic compounds. However, descriptions of a number of models are included in the
       Framework with little or no information presented on requirements forrl) adapting
       existing models for metals applications, 2) developing new metals-specific models for
       risk assessment, 3) establishing data requirements for model calibration, or 4)
       determining suitable techniques for estimating parameter values (and associated
       uncertainties). The SAB finds that further guidance will need to be developed in this
       area.

    •  Removing Section on Metal Research Needs.  The SAB feels strongly that the
       identification of research needs should not be within the scope of the current Framework.
       The SAB notes that in the Framework there has not been a thorough review of all
       research areas and it is not appropriate in the given context to highlight and identify
       specific research needs for the future. Therefore, the SAB recommends that the research
       needs section (Section 5) of the Framework be removed. A separate, follow-up
       document identifying  and prioritizing research needs would be helpful if it were done in a
       comprehensive manner. The Framework could refer to this separate document. Research
       recommendations provided by the SAB in this report could be included in the separate
       research needs document.

      While these recommendations call for substantial revision of the Framework, the SAB
commends EPA for initiating the development of a much-needed risk assessment framework for
toxic metals and metalloids and strongly urges EPA to continue developing the document. The
SAB provides specific comments and recommendations to improve the Framework.  Given the
large number of recommendations, the SAB identified those that could be completed in the
"short-term" (i.e., within six months), and those that require a more "long-term" effort. This
determination is based on the subjective judgment of the SAB.  The next version of the
Framework is not expected to fully address the long-term recommendations but should
acknowledge them.
                                          XV)

-------
2.
  Review of EPA's Draft Framework for Inorganic Metals Risk Assessment

      A Report by the Science Advisory Board Metals Risk Assessment
                        Framework Review Panel

INTRODUCTION
   The U.S. Environmental Protection Agency (EPA) Science Advisory Board (SAB) Metals
Risk Assessment Framework Review Panel has reviewed EPA's draft document Framework for
Inorganic Metals Risk Assessment (Framework).  This report transmits the SAB's comments and
recommendations. Many EPA programs face decisions on whether and how to regulate metals.
These decisions range from setting standards or permits for environmental releases, to
establishing safe levels in different environmental media, to setting priorities for programmatic
or voluntary efforts.  EPA developed the draft Framework for Inorganic Metals Risk Assessment
to supplement previous Agency guidance for use in site-specific risk assessments, criteria
derivation, and other similar Agency activities related to metals.

   EPA has followed a stepwise process to develop the draft Framework. A Metals Action Plan
(MAP) was first developed to establish a process for application of scientific principles to metals
risk assessment.  In September 2002, the SAB reviewed the MAP and provided comments to
EPA (U.S. EPA SAB, 2002).  EPA then developed metals issue papers addressing the following
topics: environmental chemistry of metals, biovailability and bioaccumulation of metals, metal
exposure assessment, human health effects, and ecological effects (U.S. EPA, 2004).  The draft
Framework document was then completed, and a peer consultation workshop was held in July
2004 to seek input on the document from scientists in the field of metals risk assessment. The
Framework document was then revised and provided to the SAB for review. The draft
Framework document contains five sections: 1) an introduction  describing the purpose, audience,
and scope of the document; 2) a problem formulation section setting forth principles of metals
risk assessment and providing a conceptual model for metals assessments; 3) a section providing
recommendations that should be considered when conducting assessments of the  ecological or
human health risks associated with metals exposures; 4) a section on metal specific topics and
methods in environmental chemistry, human health exposure pathway analysis, human health
effects, ecological exposure pathway analysis, and characterization of ecological  effects; and 5) a
section identifying metals research needs.

   The SAB commends EPA for recognizing the need to carefully analyze the differences
between metals and organic chemicals in site specific and national risk assessments.
Specifically, the SAB congratulates EPA for initiating the development of a risk assessment
framework for metals that covers a broad spectrum of topics related to human health and
ecological risk concerns from exposure to  toxic metals and metalloids. The SAB's comments
are directed to the EPA to help develop a strong final document that will help guide Agency risk
assessors for a  number of years into the  future.

-------
 3.      CHARGE TO THE JREVIEW PANEL

   EPA sought comment from the SAB on the scientific soundness of the Framework's synthesis
 and representation of the state of the science.  Specifically, EPA sought comment on: the overall
 objectivity and utility of the recommendations and supporting tools, methods, and models to its
 primary audience (EPA risk assessors and the public), and whether there are any additional
 research needs that warrant inclusion or further discussion in the Framework. EPA defined
 objectivity as: "a focus on whether the disseminated information is being presented in an
 accurate, clear, complete, and unbiased manner, and as a matter of substance, is accurate,
 reliable, and unbiased." EPA defined utility as: "the usefulness of the information to its  intended
 users, including the public."  The EPA gave the following eighteen charge questions to the SAB
 panel.

 Question 1: Section I - Framework Scope and Assessment Categories

 1.1    Please comment on the overall framework scope and whether it is sufficiently
       encompassing to allow for the consideration of the broad spectrum of physical and
       chemical properties, exposures, and effects among inorganic metals and metal
       compounds.

 1.2    The context of the regulatory application (e.g., contaminated site clean-up, national
       regulation, or programmatic decision)  is a major factor in determining the type of
       analysis that is appropriate for a particular assessment.  The framework identifies three
       general categories of assessments, including site-specific assessments, national scale
       assessments, and national ranking and  categorization. With the understanding that
       screening and detailed assessments occur within the assessment categories, please
       comment on the utility of these categories in setting the context for discussion of metals
       assessment.

 Question 2: Section 2 - Problem Formulation, Metals Principles,  and Conceptual Model

2.1    Please comment on whether the discussion of inorganic metals assessment principles is
       clearly articulated, objective, as defined above, and has utility.

2.2    Please comment on how well the conceptual model presents key metal processes  and
       whether (or not) it is complete.

Question 3: Sections 3, 4, and 5 — Recommendations, Tools/Methods, and Research Needs

3.1     Please comment on  how well the recommendations under Section 3 are supported by the
       detailed information in Section 4. Are there other recommendations that should be
       included?  Are there any  inorganic metals or metal compounds for which any of the
       recommendations would  not apply?

       Note:  Recommendations pertaining to environmental chemistry are distributed
       throughout Section 3,  particularly under Sections 3.2.1 and 3.3.1 presenting
       recommendations on environmental fate and transport.

-------
3.2    Please comment on the objectivity and utility of the data, tools, and methods discussed in
       Section 4. Identify any scientific or technical inaccuracies, or any emerging areas or
       innovative applications of current knowledge that may have been overlooked or warrant a
       better discussion of uncertainty, including areas needing further research.

3.3    Please comment on the state of the science (i.e., data, tools and methods) to address
       inorganic metals speciation in all environmental compartments for any given  inorganic
       metal from the point of environmental release to the point of toxic activity as  discussed in
       the document. Please comment on whether the framework identifies appropriate research
       needs to overcome any limitations in the state of the science. Please address these
       questions separately for each of the three types of assessments presented (i.e., site-
       specific, national level, and ranking and categorization.)

3.4.    In an earlier draft of the framework, EPA had included three Summary Recommendation
       Tables in Section 3 on human health, aquatic, and terrestrial risk assessment,  covering the
       three general  assessment categories (i.e., site-specific, national level, and ranking and
       categorization). An example of this table is included as Appendix A in the draft provided
       to the SAB. To minimize confusion for users of the framework, the initial idea behind
       the recommendations and adjoining table was to have concise recommendations on the
       science, followed by a separate accounting of how these recommendations could then be
       applied to the different assessment categories.  Reviews have been mixed on the utility of
       these tables as a sufficient communication tool. Please comment on whether  tables of
       this type would be useful for inclusion in the final version of the framework.  Does the
       panel have alternative suggestions for effectively communicating how the
       recommendations can be considered for each of the three assessment levels?

Environmental Chemistry (Sections 3.3.1, 4.1)

3.5    Please comment on the objectivity of the Hard Soft Acid Base concept to applications of
       stability of metal complexes in toxicity assessments.  See Section 4.1.2.

3.6    Please comment on the objectivity of the atmospheric metal chemistry discussion and  its
       application to exposure assessments.  See Sections  3.3.1.1 and 4.1.7.

3.7    Please comment on the objectivity of the metal chemistry and environmental  parameters
       incorporated in the various metal surface complexation and partition coefficient models
       and their applications to  exposure assessments. See Sections 3.3.1.2 and 4.1.4.1.

Human Exposure and Health Effects (Sections 3.1, 4.2, 4.3)

3.8    Please comment on the objectivity of the discussion and recommendations on natural
       background of metals. See Sections 3.1.2.1 and 4.2.2.1.

3.9    Please comment on the objectivity of the discussion of essentiality versus toxicity,
       including the relationship between Recommended Daily Intakes (RDAs) and thresholds
       such as Reference Doses (RfDs) and Reference Concentrations (RfCs). See Sections 3.1,
       4.3.2, and 4.3.3.

-------
 3.10   Please comment on the objectivity of the discussion and recommendations presented for
       assessing toxicity of mixtures, including how to assess additivity versus departure from
       additivity.  See Sections 3.1.3.4 and 4.3.6.

 Ecological Exposure and Effects (Sections 3.2, 3.3, 4.4, 4.5)

 3.11   Please comment on the objectivity of the discussion and recommendations concerning
       natural background, bioavailability, bioaccumulation, biomagnification, and trophic
       transfer in both aquatic and terrestrial environments. See Sections 3.2.2 to 3.2.4, 3.3.2,
       4.4.3, 4.5.4, and 4.5.6 to 4.5.9.

 3.12   Please comment on the objectivity of the framework statement that the latest scientific data
       on bioaccumulation do not currently support the use of bioconcentration factor (BCF) and
       bioaccumulation factor (BAF) values as generic threshold criteria for hazard classification
       of inorganic metals (see recommendation on page 3-17,  lines 27-29 of the document). By
       this, the framework means that various assumptions underlying the BCF/BAF approach,
       including the independence of BCF/BAF with exposure  concentration and the
       proportionality of hazard with increasing BCF/BAF do not hold true for the vast majority
       of inorganic metals assessed.  Please comment on the framework's acknowledgement that
       the appropriate use of BCFs/BAFs to evaluate metal bioaccumulation, including the degree
       to which BCFs/BAFs are dependent on exposure concentrations, needs to consider
       information on bioaccessibility, bioavailability, essentiality, acclimation/adaptation,
       regulation of metals (uptake and internal distribution), detoxification and storage,
       dependence on exposure concentration, and background accumulation. While the ability to
       quantitatively address ail these factors  may be limited at the present time, the framework
       states that their potential impacts should at  least be qualitatively addressed.  See Sections
       3.2.4, 3.3.2.5, and 4.5.3.

3.13   Given the variety of organism responses to  inorganic metals exposure, based on factors
       such as bioaccessibility, bioavailabih'ty, essentiality, uptake/excretion mechanisms, and
       internal storage/regulation, as described in Section 3.2.4, the framework states that
       BAFs/BCFs should be derived using mathematical relationships that represent the
       concentration in the organism or tissue as a function of the bioavailable concentration in
       the exposure medium/media for each set of exposure conditions. Please comment on
       whether this is the  besl approach based on the current state of the science or  if there are
       alternative approaches that are more appropriate that can be routinely applied. See
       Sections 3.2.4, 3.3.2.5, and 4.5.8.

3.14   Please comment on the objectivity of the information and recommendations pertaining to
       the use of the simultaneously extracted metals-acid volatile sulfides (SEM-AVS)
       approach and the biotic ligand (BLM)  model.  Are additional recommendations
       warranted?  If yes, what are they?  See Sections 3.2.6, 4.4.2.3, and 4.5.10.

4.     REVIEW PROCESS

   To establish the Metals Risk Assessment Framework Review Panel, the EPA Science
Advisory Board Staff Office published a Federal Register notice requesting nominations and

-------
identified a subset of nominees for consideration as panelists.  The final panel was selected after
requesting public comments on the nominees and further evaluating them against EPA Science
Advisory Board selection criteria. The members of the review panel included scientists with
expertise in: the environmental chemistry of metals, environmental fate and transport of metals,
bioavailability of metals, routes of exposure of aquatic and terrestrial species to metals, routes of
human exposure to metals, human health effects of exposure to metals, and ecological effects of
exposure to metals.

    The SAB review was conducted by a public teleconference and a two and one half day
public face-to-face meeting of the SAB Panel. During the public conference call, EPA answered
questions from the Panel about the draft Framework and the review charge. At the public
meeting, the Panel heard presentations from EPA on the Framework and deliberated on the
charge questions.  The Panel met in the following three working groups to develop responses to
the charge questions: 1) Environmental Chemistry/Fate and Transport, 2) Human Exposure and
Health Effects, and 3) Ecological Exposure and Effects/Bioaccumulation.  Responses of the three
working groups were integrated by the Panel to develop the final SAB report.
5.
OVERARCHING COMMENTS AND RECOMMENDATIONS
   The SAB provides a number of broad overarching comments and recommendations to
improve the Framework. The SAB notes that the draft Framework is an ambitious attempt to
survey the major issues involved in the assessment of human health and ecological effects of
metals and metalloids. In this regard, the SAB believes that the following major issues within
the Framework document need to be addressed before the document is published in final form in
order to make it of more current and long-term value to EPA.

   The SAB recommends substantial revision of the Framework to reorganize the document,
include additions and corrections, and  remove redundancies as detailed in the responses to the
charge questions below.  Because of the scope of recommended revisions, the SAB believes the
revised Framework would benefit from a second external peer review. The SAB is willing to
provide such a review. As discussed below, the SAB finds the Human Health section of the
Framework, in particular, to be incomplete and in need of major revision. Some critical
references are missing, a number of the references cited in Section 4 are outdated, and more
recent references should be included. The ecological subsections of the Framework should  more
fully reflect the state of the science (i.e., they leave the perception of not being objective). The
bioaccumulation and bioavailability sections need to treat the routes of exposure (diet and
dissolved metals) in an integrated fashion.  This could be accomplished by organizing the
discussion around the bioavailability conceptual mode!.  The toxicity testing  section needs to
discuss uncertainties of particular importance to metals: the lack of dietary exposure of test
animals to metals in laboratory toxicity tests is a good example.  The discussion of
simultaneously extracted metals-acid volatile sulfides (SEM-AVS) does not adequately address
the limitations of the approach (e.g., bioavailability from oxidized sediments). As noted in the
SAB review of EPA's Integrated Approach to Metals Assessment in Surface  Waters and
Sediments, SEM-AVS is of little use in oxidi/ed environments or those where sediments are
periodically rcsuspended (EPA >SAB. 2000a). Similarly, discussions of the biotic ligand model
(BLM) do not adequately describe its limitations or the early stage of BLM development.

-------
 Finally, other approaches such as the National Oceanic and Atmospheric Administration's
 (NOAA) empirically-derived effects range median (ERM) and effects range low (ERL) approach
 (Long & Morgan, 1990; 1991) should be included in the discussions.

   The following overarching comments and recommendations are discussed in more detail in
 the responses to the charge questions below.

 Presentation

   The SAB finds that the overall clarity of expression, precision of wording, and balance in
 coverage among topics in the Framework must be greatly improved. Many of the SAB's
 comments below focus on the main technical issues that need to be addressed specifically.
 However, the SAB finds that sections of the current Framework are unclear and disorganized and
 that revision is needed to develop a document that is of high quality.

 Title

   The title of the Framework is awkward: metals are inorganic by definition, and thus the use of
 the adjective "inorganic" in front of metals is redundant. Although the SAB realizes that the
 adjective "inorganic" was probably used in the original title to exclude organometallics,
 especially methylmercury, it detracts from the clarity of the title.  A better title would be
 "Framework for Assessment of Risk of Metals and Metalloids in the Environment."  At the
 beginning of the Framework document, EPA should clarify the rationale for specifically
 excluding organometallic compounds from the document while including natural transformation
 pathways that form organometallic compounds.

 Purpose

   The SAB finds that a major weakness in the current version of the Framework is the lack of
 consistency in identity. At times, the Framework provides background information on the state
 of the science and general recommendations of "basic principles" that need to be considered for
 risk assessments of metals. At other times, the report appears to serve as a practical guide for
 risk assessors, offering specific recommendations of methods and tools (often with insufficient
justification for the specific selection). This dual nature of the report stems largely from its
 intended purpose (as stated on pages 1-1 and 1-2) to serve as a "statement  of policy" while at the
 same time "provide recommendations and foster consistent application" across EPA.  The SAB
 recommends that the purpose of the Framework be reviewed and that the document be revised
 accordingly to remove any confusion in its intended purpose.  If the document is to serve as both
 a framework and practical guide for risk assessors, the recommendations and guidance in the
 document should be balanced and organized consistently with this dual purpose in mind. EPA
 should carefully differentiate material that is presented as "the framework  for assessment,"
 "examples to illustrate and clarify framework issues," and "specific instructions."  In addition, all
 recommendations in the Framework should be carefully reviewed and revised to ensure that they
 are consistent with its intended purpose. As such, the recommendations should focus on the key
 issues that need to be considered in  metals evaluations. Specific methods and tools should be
 cited accordingly to highlight the current state of the science and to serve as examples.  EPA,

-------
however, should refrain from making final recommendations of specific methods and tools until
a full evaluation of the strengths and weaknesses of each method and tool is performed.

Critical Evaluation of Supporting Information

   The SAB commends EPA for providing fairly comprehensive coverage of available tools for
risk assessment and methods for metals analyses. In many instances however, critical
evaluations of the tools and methods are not provided and the justification for many
recommendations is not clear. The SAB therefore recommends that more information be
presented on the strengths, weaknesses, and limitations of the various methods and tools. Where
appropriate., comparative assessment of competing approaches  should be provided.

Tiered Recommendations in the Framework

   The SAB recommends that the recommendations in the Framework be tiered, with the most
critical general overarching recommendations (those with the greatest impact) presented first,
followed by specific recommendations that would be of value to the assessor. This would help
focus the different sections of the Framework to ensure that the most important issues are
addressed.

Illustrative Examples

   Illustrative examples would be useful throughout the document. Examples of how certain
recommendations might be implemented would greatly improve the utility of the document.
Identification of important metal sources such as accumulation from coal mining, chromium
from plating facilities, silver from photographic facilities and atmospheric deposition of mercury
to watersheds might provide an indication of the diverse range  of sources that should be
examined.

Discussion of Uncertainties and Data Quality

   Discussions  of uncertainties of the  tools, methods and data are generally lacking and
inconsistent throughout the document. Data quality  is a large concern for metals, particularly
measurement of dissolved metals.  Historic data must be considered with a critical eye, as the
data were often generated before clean-room and trace-level measurement techniques were
adopted.  The need to critically consider data quality should be explicitly stated throughout the
document wherever the use of analytical data is discussed.

Terminology and Additions to the Glossary

   As discussed in the detailed responses to the charge questions, and in the recommendation
concerning the definition of bioaccumulation below, the SAB recommends revision of several
definitions in the glossary to make them consistent with current science and reduce confusion to
the reader.

-------
 Use of the term "Bioaccumulation " versus "Accumulation " to Describe Metals Concentrations

    It is the opinion of the SAB that there should not be a distinction in the Framework between
 the term "bioaccumulation" to describe metal concentration in aquatic and terrestrial organisms
 and the term "accumulation" of metals for humans.  This is not an accepted distinction in the
 scientific community. In humans as in other terrestrial animals, the steady-state body burden of
 many metals is under homeostatic control that balances intake and excretion.  However, for
 certain metal compounds bioaccumulation can occur, which can be defined as either a persistent
 increase in individual steady-state levels  that is correlated with higher prior exposure, and/or a
 progressive increase in body burden as a  function of exposure time or age, that is above normal
 steady-state levels and which may involve selective bioaccumulation of the metal in certain
 tissues.

   The SAB believes it imporant to recognize that some metals do bioaccumulate in the tissues
 of humans and that this bioaccumulation  is related to their toxicity.  The rate at which this
 process occurs depends upon the balance between the accumulation and elimination of the metal
 in the tissues of concern and, thus, is dependent upon the concentration of the exposure dose and
 the frequency  of exposure. Pharmacokinetic models can be used to estimate the extent to which
 metals bioaccumulate in tissues.  The SAB  recommends that the definition of the term
 "bioaccumulation" in the glossary of this document be modified to read as follows:

    Bioaccumulation: The net accumulation of a metal in a tissue of interest or the whole
    organism that results from exposure from all environmental sources, including air, water,
    solid phases (i.e. soil, sediment) and diet, and that represents a net balance of uptake versus
    elimination of the metal.

Metal-specific Reference Values (RfD/RfC) and/or Cancer Potency Factors

   The role of the human risk assessor is  to appropriately integrate the reference values and
potency factors with the exposure assessment.  Thus the risk assessor needs an understanding of
the toxicological endpoints and mechanisms of action that underlie the derivation of these values
to ensure that,  for example, the appropriate population and life stages are addressed, appropriate
dietary aspects are taken into consideration, and the appropriate exposure pathways are
considered. For metals, frequency and duration of exposure, as well as exposure concentrations,
are important parameters to be considered for accurate dose assessments. The discussion in the
Framework should focus on advising human health risk assessors on how to take these
considerations into account in constructing  the risk assessment. The SAB recommends that, in
introducing the Human Health  Effects section, EPA should set the context by explaining that
human health risk assessors start their analysis with a metal-specific reference value (RfD/RfC)
and/or cancer potency factor that has been 'developed through a process separate from the risk
assessment.

Background Versus Ambient Concentration

   The concept of background  levels of metals as described in the Framework is not clearly
defined.  The SAB therefore recommends using the term "ambient"  or "ambient levels" rather

-------
than background. The term background is often incorrectly assumed to connote "natural" and
therefore "safe" or of no significant human or ecological health concern.  However, ambient
levels can vary, or can be inherently high enough to represent a  potential health concern in and
of themselves. They can also represent a total level from a combination of natural and
anthropogenic sources, some of which may be historical or unknown. For metals in particular,
the concept of background levels as described in the Framework document is complicated by
several factors, which include the sometimes highly variable natural levels of metals in soils,
sediments, air and water, various historical anthropogenic sources or activities, and air deposition
from distant anthropogenic sources.  This is also discussed in detail in the response to charge
question 3.8 below.  The SAB acknowledges, however, that the term "background" exposure has
been used in human health dose response assessment. This term, referring to both the natural
and anthropogenic sources of a chemical under  assessment, has been used in various publications
on dose-response modeling over the past 25 years or more.

Chemical Speciation

   Among risk assessors and scientists working on metals, the concept of "chemical species" and
"chemical speciation" is fundamental. In the Framework, there are certain instances where the
terms are used incorrectly. This is discussed in the response to charge question 3.3 below. The
SAB recommends that, in addition to correcting these instances, the speciation concept be
introduced in the environmental chemistry part  of Section 2, specifically  in the "environmental
chemistry" principles section, and in the environmental chemistry part of Section 4. Appendix B
of this SAB report contains text that is adapted  from recent IUPAC recommendations
(Templeton et al., 2000). The SAB believes that this material would serve as a suitable starting
point for discussions in Section 4 of the Framework.

   The SAB also recommends  that greater care be taken in distinguishing general descriptions of
solid-water "partitioning" processes and the very specific term "partition coefficient."  In this
context, "partitioning" refers to a general set of processes that controls the distribution of metal
among dissolved and solid phases, whereas "partition coefficient" is one  specific descriptor of
the empirical distribution which is based on the ratio of solid phase to dissolved metal.

Balance of Coverage - Metal Specialion

   The SAB commends the EPA for emphasizing approaches that employ a relatively
sophisticated understanding of metal  speciation in the context of metals risk assessment. While
there is an adequate discussion in the Framework of the use of models to estimate metal
speciation in water, soil, and sediments, there is insufficient discussion of analytical tools to
measure the speciation of a metal. A fuller description of the tools that are currently available to
quantify metal speciation in environmental samples, including the strengths and weaknesses of
each technique, would be of great benefit to a risk assessor in determining the form and potential
effects of metal contamination  at a given site, and which tools are most appropriate for a given
assessment.

-------
 Metals Mixtures

   The SAB notes that in virtually all settings, individual metals exist as components of
 mixtures. Even in their natural settings, metals of concern to a risk assessor are typically
 mingled with other metals. When the question of risk is posed from the standpoint of pollution
 episodes, the principle still holds; that is, metals are usually presented to ecological receptors and
 to humans as a mixture with ether metals and/or organics. In all instances and settings, then, the
 assessor must be aware of the additional materials present in that particular environment when a
 metal is studied as a potentially hazardous pollutant. These "mixed exposures"  can have
 dramatic effects on the toxic potential of the metal.

 Mimicry

   The SAB notes that structural similarities of metals, such as similar ionic radii, may result in
 competition for essential receptors, thus, disrupting normal functions. Examples may include
 chromate substituting for sulfate or phosphate, Pb replacing Ca or Zn, and Cd substituting for Zn
 or Ca on important regulatory proteins or enzymes. The degree to which these ionic
 substitutions occur in target cell populations is dependent upon a number of factors including
 cellular uptake/excretion of toxic metals, intracellular complexations with metal-binding proteins
 such as metallothionein  or lead-binding proteins and sequestration in lysosomes or inclusion
 bodies. In this  regard, the limited discussion in the Framework of metal-binding proteins should
 be expanded to include more recent references on all of these potential intracellular metal
 sequestration depots since they will determine the extent to which molecular/ionic mimicry
 actually occurs in vivo (see response to charge question 3.10).

 Balance of Coverage — Data Collection

   The SAB finds that the Framework contains insufficient information on appropriate
 parameters for data collection. Recommendations and supporting information should be
 presented on the types of field data that are needed (including metal speciation and
 concentrations, and related system parameters such as pH, redox conditions, organic carbon
 concentrations, iron concentrations, acid volatile sulfides, etc.), and on the appropriate time and
 space scales for data collection. Revised procedures and processes that are needed to evaluate
 the adequacy and quality of tbs data being used for the metals risk assessment should be
 discussed.

 Biogeochemistry

   The SAB notes  that a key difference in the  fate and transport of metals as compared to
 organic compounds is in the  relationship of metals to biogeochemical cycles.  For organic
 compounds, the coupling to natural biogeochemical cycles is essentially unidirectional from the
 risk assessment perspective (i.e., the major biogeochemical cycles affect the fate and transport of
organic compounds, but not vice versa).  Metals  interact with the cycles of more elements
(especially sulfur and  other metals) than organic compounds. In addition, metals can be limiting
nutrients or toxicants to organisms that drive the major biogeochemical cycles (e.g., higher
plants, phytoplankton, bacteria). The SAB finds that the role of metal biogeochemical cycling is
                                            10


-------
not adequately addressed in the conceptual model for the risk assessment framework, and in
subsequent sections of the report (see response to charge question 2.2).

Modeling

   The SAB notes that the  Framework accurately reflects the fact that modeling the
environmental fate and transport of metals differs in significant ways from modeling organic
compounds. However, descriptions of a number of models are included in the Framework with
little or no information presented on requirements for: adapting existing models for metals
applications, developing new metals-specific models for risk assessment, establishing data
requirements for model calibration, or determining suitable techniques for estimating parameter
values (and associated uncertainties). Further guidance will need to be developed.

Overarching Comments on Specific Sections of the Framework

    •   The "principles" provided in Section 2 of the Framework are not fundamental principles.
       The term, "principles," should therefore be replaced with a more appropriate term such as
       "factors" or "key issues."  The SAB also finds a  lack of uniformity in the quality and/or
       clarity  of writing among the parts of Section 2. It is noted that the report of the SAB's
       2002 Metals Assessment Plan (MAP) review (EPA Science Advisory Board, 2002)
       addressed many of the same issues. It is therefore recommended that the SAB MAP
       report be revisited prior to revision of Section 2 in order to improve the quality and
       clarity  of the writing.

    •   Section 3 of the Framework should be reorganized to provide more comparability among
       the parts  of the section. Recommendations should be highlighted by minimizing textual
       justification and cross-referencing justification to Section 4.

    •  The recommendations in Section 3 of the Framework should be  rewritten to clearly
       express them as recommendations (rather than statements).

    •  The number of recommendations in Section 3 of the Framework should be reduced by
       omitting  statements and condensing similar or redundant recommendations.
       Recommendations  should also be organized by importance or specificity.

    •  Revised recommendations in Section 3 of the Framework should not be prescriptive, but
       suggest options or examples.

    •  Tables such as those provided in A-2 of the Framework should be included in an
       appendix.  Recommendations for improvements  to the tables are provided below  in the
       response to charge question 3.4.

    •  Section 4 of the Framework should be reorganized to mirror the organizational structure
       used in Section 3.

-------
    •  As indicated in the response to charge question 1.1 below, Section 5 of the Framework,
       "Research Needs", should be removed from the document because the research needs are
       not supported with interpretative text. A separate, follow-up document identifying and
       prioritizing research needs would be helpful if it were done in a comprehensive manner.

 6.     RESPONSE TO THE CHARGE QUESTIONS

   In the responses to each of the charge questions below, the SAB provides a section devoted to
 detailed comments followed by key recommendations to EPA.  Many of the SAB
 recommendations take the form of suggested changes to wording to improve clarity.  These can
 be found in Appendix C. In addition, the SAB has identified each recommendation herein to be
 either "short-term" or "long-term."  Recommendations that, in the opinion of the SAB, can be
 addressed in less than six months are identified as "short-term" recommendations.
 Recommendations requiring more time to implement are identified as "long-term"
 recommendations. The determination of whether recommendations are long or short term is
 based on the subjective judgment of the SAB. The next version of the Framework  is not
 expected to fully address the long-term recommendations but should acknowledge  them.

 6.1.1    Charge Question 1.1. Please comment on the overall framework scope and
         whether it is sufficiently encompassing to allow for the consideration  of the broad
         spectrum of physical and chemical properties, exposures, and effects  among
         inorganic metals and metal compounds.

 6.1.1.1   Comments in Response to Charge Question 1.1

   The SAB generally finds that the overall Framework scope is sufficiently broad  and provides
 an appropriate level of flexibility in addressing issues of concern. However, the SAB has
 identified a number of specific issues that should be addressed.  The SAB finds  that the
 following four specific issues deserve attention in answering charge question 1.1.

 Balance Between Science and Guidance

   As noted above, the Framework document has features of both a state-of-science document
 and a technical guidance document. The SAB recommends  that the Framework be  reviewed and
 revised to remove any confusion in its intended purpose. If the  document is to serve as both a
 framework and practical guide for risk assessors, the recommendations and guidance in the
document should be balanced and organized consistently with this dual purpose in mind. EPA
should carefully differentiate material that is presented as  "the framework for assessment,"
"examples to illustrate and clarify framework issues," and "specific instructions." It is also
 important to note that critical evaluations are  needed to ensure that the Framework does not
prescribe the use of specific methods or tools for risk assessment that may become obsolete over
time.

 Treating Human and Ecological Health Risk Assessment in One Document

   The SAB agrees that both human and ecosystem  health risk assessment need to be in one
framework document since the uniqueness of metals compared to organic compounds is germane
                                          12

-------
to both. However, the document needs to achieve better balance in quality and depth of coverage
in the sections on human and ecosystem health. Better integration of the human health and
ecological health sections with the environmental chemistry section is also needed.

Expanding and Clarifying the Definition of Metals

   The SAB feels that the use of the term "metals and metal compounds" is confusing and does
not accurately capture the types of metals EPA intends to cover in the document.  The SAB
recommends that the introduction section of the Framework provide a definition and
nomenclature that is inclusive of metals that do not behave like organic compounds but also
delineates the groups and classes of metals covered by this document, including metalloids.

Removing Section on Metal Research Needs

   It is the opinion of the SAB that the identification of research needs should not be within the
scope of the current Framework. There has not been a thorough review of all research areas, and
it is not appropriate in the given context to highlight and identify specific research needs for the
future. Therefore, the SAB recommends that the research needs section (Section  5) of the
Framework be removed.

6.1.1.2   Key Recommendations in Response to Charge Question 1.1

Short-term

1. The purpose of the Framework should be more clearly defined, and the document should be
reviewed and revised to remove any confusion in its intended purpose.

2. The Framework should be reviewed to ensure that it does not prescribe specific methods or
tools for risk assessment that may become obsolete over time.

3. The introduction section of the Framework should provide a definition and nomenclature that
is inclusive of metals that do not behave like organic compounds, but also delineates the groups
and classes of metals covered by the document, including metalloids.

4. The research needs section (Section 5) should be removed from the Framework because the
document does not contain a thorough review of all research areas.

Long-term

5.  The Framework should be revised  to  achieve better balance in quality and depth of coverage
in the sections on human and ecological  health.  Better integration of the human health and
ecological health sections within the environmental chemistry section is also needed.

6.1.2    Charge Question 1.2. The context of the regulatory application (e.g., site specific
         contaminated site clean-up, national regulation, or programmatic decision) is a
         major factor in determining  the type of analysis that is appropriate for a
                                           13

-------
         particular assessment. The framework identifies three general categories of
         assessments, including site-specific assessments, national scale assessments, and
         national ranking and categorization. With the understanding that screening and
         detailed assessments occur within the assessment categories, please comment on
         the utility of these categories in setting the context for discussion of metals
         assessment.

6.1.2.1   Comments in Response to Charge Question 1.2

   In general, the SAB finds that the risk assessment categories listed in the Framework are an
appropriate context to cast the relevant issues of metals in comparison to organic compounds.
The Framework document needs to consider the important properties of metals in these
regulatory contexts.

   The SAB, however, recommends that the scope of the categories be more clearly defined at
the beginning of the document. For example, the SAB believes that the three categories
delineated in the document m&y actually represent five different aspects of assessment (national
screening level assessment, national ranking assessment, national complex assessment, site scale
screening assessment and site scale complex assessment).  Examples of these kinds of
assessments are provided below. Definition of these five assessment categories is necessary
because in each category the assessment differs in scope and complexity. Under national
ranking and categorization, single metal  properties or regional site features can be used.
Similarly, at the national level assessment, a single parameter can be utilized or the assessment
can incorporate site-specific information. At the site-specific assessment level, however, the
approach is more focused. Examples of the types of risk assessment that span the range of
complexities referred to above include national level  that can be: 1) screening (e.g., comparing
ambient water concentrations to water quality criteria), 2) ranking (e.g., a contaminant candidate
list for the Safe Water Drinking Act); or 3) complex (e.g., criteria documents).  They can also
include more site-specific screening such as that required prior to completing an environmental
impact statement; and site-specific complex assessments such as those required for Superfund.

   The SAB feels that the sections in the Framework following the introduction largely concern
site-specific assessment issues. The SAB therefore recommends that the document be edited to
represent more balance among the different  types of assessment. In addition, the document
should include focused discussions and mapping to relevant issues at each level of assessment.

6.1.2.2   Key Recommendations in Response to Charge Question 1.2

Short-term

1.  The scope of the general categories of assessments should be more clearly defined  at the
beginning of the document. Examples of the types of risk assessments that spaa the range  of
complexities should be provided.

2.  The SAB finds that the sections in the Framework following the introduction largely concern
site specific assessment issues, and recommends that  the document be edited to represent more
balance among the different types of assessment.
                                          14
J

-------
6.2.1     Charge Question 2.1. Please comment on whether the discussion of inorganic
         metals assessment principles is clearly articulated, objective, as defined above,
         and has utility.

6.2.1.1   Comments in Response to Charge Question 2.1

   Section 2 of the Framework is entitled Problem Formulation and Principles. This suggests
that Section 2 will provide a concise overview of the Framework. The SAB finds that some
changes are needed to make this view of the Framework consistent with recommendations in
Section 3 and the detail in both Section 4 and EPA's Papers Addressing Scientific Issues in the
Risk Assessment of Metals (Issue Papers) (EPA, 2004). The SAB also finds a lack of uniformity
in the quality and/or clarity of writing among the subsections in Section 2 of the Framework. In
addition, the SAB finds that Section 2 of the Framework has an imbalance of coverage among
the principles considered. The SAB provides recommendations to address these concerns and
improve the utility, objectivity, and clarity of the document.

Articulation of the Inorganic Metals Assessment Principles

    A primary issue that arises concerning the utility of the material in Section 2  of the
Framework is applicability of the material at local, regional, and national scale risk assessments.
It is the judgment of the SAB that most of the detailed material in Section 2, and indeed
throughout the Framework, is relevant to site-specific risk assessment. However, the general
descriptions of the "principles" are relevant to larger-scale risk assessments as well as site-
specific assessments.

   The topics listed in Section 2 of the Framework are not principles but rather factors to be
considered. For example, bioaccumulation is a process; the relevant principle is activity. The
SAB recommends that the terminology in the Framework be changed.  It is recommended that
EPA drop use of the word "principles" and instead use "factors to be considered" or "factors."
The SAB supports the inclusion of the "key  questions" listed under several, but not all, of the
factors in the Framework. It is recommended that "key questions" be listed in the front of the
subsections for all factors included.  This will result in parallel construction and help justify the
selection of metal-unique topics to focus on  in Section 3.  The key questions  should identify why
factors are  important and uniquely need to be considered for metal risk assessments.

   The SAB finds a lack of uniformity in the quality and/or clarity of writing among the
subsections in Section 2 of the Framework.  It is noted that the report of the SAB's 2002 Metals
Assessment Plan (MAP) review (EPA Science Advisory Board, 2002) addressed  many of the
same issues. It is therefore recommended that SAB MAP report be revisited prior to revision of
Section 2 in order to improve the quality and clarity of the writing in some subsections. Some of
the material in the SAB MAP report may be used in Section 2.

   The SAB also finds that Section 2 of the Framework has an imbalance of coverage among the
factors considered.  For example, the subsections on environmental chemistry and toxicity
testing are very brief and other important processes such as dietary and or food web exposure
and atmospheric transport to receptors should be discussed under route of exposure.  It is
recommended that the extent of the discussion in the subsections be reviewed and made more
                                           15

-------
uniform.  Suggestions for specific revisions in this regard are provided below (see especially the
recommendations for subsections 2.1.4,2.1.5, and 2.1.6).

   In the context of risk assessment, the factors included in the Framework comprise a fairly
complete list, but some important factors have been omitted and should be added to the text.
These are the nature and type of source, and the route of exposure.  These factors should be
added to the list in the Framework and text should be developed to a level of detail that is
consistent with the other factors presented. While the two factors noted above are relevant for all
contaminants, there are unique aspects of metals sources and routes of exposure that a risk
assessor will have to address. Two important processes that should be discussed under route of
exposure are trophic (dietary and/or food web) transfer, and atmospheric transport to receptors.
In the explanation of trophic transfer, it should be noted that the concentration in the water is not
predictive of the concentrations at the highest trophic levels. With regard to atmospheric
transport, it should be noted that most metals occur almost exclusively as particles in the
atmosphere, and this  affects how exposure occurs and the types of effects exerted on receptors.

Objectivity and Utility of Inorganic Metals Assessment Principles

   Section 2 of EPA's Framework document provides an overview of the risk assessment
framework for metals, including the conceptual model representing the various components of
the process and their interlinkages. The SAB finds Section 2 to be of high utility for
understanding the context of the recommendations in Section 3 and the importance of the
detailed process component descriptions in Section 4. However, the SAB provides the following
recommendations to improve the utility, objectivity, and clarity of the document.

    •  The introductory paragraphs of Section 2 on page 2-1 of'the Framework emphasize the
       need for risk assessments at scales ranging from site specific to national. It would be
       useful to note the risk assessment factors that are unique to  metals. It would be helpful
       to clearly discuss how the complex properties and reactivity of metals present unique
       challenges in risk assessment.

    •  The terms used to describe the various factors introduced in Section 2 also need to be
       carefully defined. For example, the term "essentiality" is vaguely defined in comparison
       to the level of detail in text boxes defining "background" and "bioavailability." A more
       precise definition of essentiality that should be  included in the document is, "a metal that
       participates in and is required for some basic biological process with positive
       consequences for the organism." Similarly, "bioaccumulation and bioconcentration"
       could be defined in a text box that incorporates the definitions of "bioconcentrate,"
       "bioaccumulate," and "biomagnify" that are presently in the text. A definition of trophic
       transfer should also be included in this text box. The SAB also notes that the definition
       of bioavailability given on page 2-6 of the Framework and in the glossary suggests the
       units of a rate constant in an uptake equation. This does not fit the intended definition of
       the term.

    •  The discussion of "background" in subsection 2.1.1 of the Framework includes
       references to both naturally occurring and anthropogenically-introduced metals. To
                                           16

-------
some reviewers, the subsection seemed to imply that risk assessments should focus on
rnetals present above natural system concentrations. The SAB therefore recommends
that in this subsection EPA place greater emphasis on the potential for naturally
occurring metals to pose as much risk as anthropogenic metals.  The SAB notes that
arsenic, for example, is naturally occurring but still needs to be regulated. It should be
more clearly emphasized in the Framework that background concentrations are not
necessarily acceptable concentrations.  The SAB also notes that consideration of
background is substantially different for risk assessments conducted at local, regional,
and national scales.

The SAB notes that involvement of metals in biogeochemical cycles should be
emphasized in the Framework in the discussion under the factor "environmental
chemistry."  At the ecosystem scale, metal biogeochemical cycling considerations are
different for metals than for organic compounds.  Since metals do not biodegrade, they
are recycled in the environment.  Metal cycles are often coupled with nutrient cycles.
This has important implications for risk assessment since metal contaminants may not
pose a risk in the current environmental scenario under consideration, but they may pose
a future risk if their chemistry (e.g., oxidation-reduction  conditions) changes. In this
context, there may not be any single value of "bioavailable fraction" (mentioned on page
2-2, lines 3-4 of the Framework) of a metal that applies to  its fate once discharged to the
environment.  In the environmental chemistry section, metal fate, transport, and
bioavailability should be discussed in the context of biogeochemical cycles.

The environmental chemistry section of the Framework currently focuses on speciation.
The SAB believes that additional issues should be included in this section of the
document. Other issues that involve unique considerations for metals include processes
affecting metals in sediments, and reactions that incorporate metals in organic
compounds such as  methylation.

The "bioavailability" subsection of the Framework (2.1.4) is much longer and more
detailed than the other sections. To improve the utility of this part of the Framework, the
SAB recommends the following revisions. The conceptual
bioaccessibility/bioavailability model shown in Figure 2-2 should be moved to Section 4,
as should the "bioaccessibility", and "bioavailability" sections.  The first italicized
sentence in section 2.1.5 ("Bioaccumulation and Bioconcentration") defines the
bioaccumulation issue, but the rest of ihe section appears to be a scattered set of
observations that do not help define what is unique to metals about bioaccumulation,
what is of concern with how the issue is used (the specific construct), or how it might be
used in risk assessments. The discussion should be revised to address these questions.

Subsection 2.1.6 ("Acclimation, Adaptation, and Tolerance")  is an important component
that should be linked to the discussion of essentiality in subsection 2.1.2.  Also,
subsection 2.1.6 should include the potential costs (e.g.,  genetic erosion) of the
acclimation, adaptation, and tolerance phenomena when or where they occur (some
discussion should be brought forward from Section 4), as well as their influence on
toxicity testing.
                                    17

-------
    •   The Framework addresses those assessment issues associated with inorganic compounds
        but also discusses natural transformation pathways that form organometallic compounds.
        The mixtures discussion in the Framework document focuses on metal mixtures.  The
        SAB notes however, that the document should also contain a discussion of interactions
        between metals and organic chemicals as it applies to the problem of mixtures. Mixtures
        of metals and certain organic compounds can behave additively, synergistically and/or
        antagonistically with respect to cancer risk, depending on the mixture and the context.
        There is ample evidence of this from laboratory experiments with simple mixtures (e.g.,
        arsenic and PAHs) showing a variety of complex effects not well predicted by
        knowledge of either agent alone. In addition, it would be useful to include a discussion
        indicating that metals can react with organics to form organometallic compounds, thus
        transforming a metal to a state in which its fate and risk will be governed by processes
        more relevant to organic compounds (e.g., biodegradation, partitioning  to dissolved
        organic carbon [DOC]).

6.2.1.2   Key Recommendations in Response to Charge Question  2.1

Short-term

1. EPA should drop the use of the word "principles" in the Framework and instead use "factors
to be considered" or "factors."

2. EPA should list "key questions" in the front of appropriate Framework subsections for all
factors included.

3. EPA should revisit the SAB Metals Action Plan report (EPA Science Advisory Board, 2002)
prior to revision of Section 2 of the Framework in order to improve the quality and clarity of the
writing in some subsections.

4. EPA should review the extent  of the discussion in all parts of Section 2 of the Framework and
make it more uniform. Suggestions for specific revisions in this regard are provided above (see
the recommendations for subsections 2.1.4, 2.1.5, and 2.1.6).

5. Some important factors that have been omitted from the Framework should be  discussed in
the text.  These factors are: the nature and type of source, and route of exposure.

6. Risk assessment factors unique to  metals should be identified in the text of the  Framework.
The document should discuss how the complex properties and reactivity of metals present unique
challenges in risk assessment.

7. EPA should carefully define the terms used to describe various factors introduced in Section 2
of the Framework.

8. EPA should place greater emphasis in the Framework on the potential for naturally occurring
metals to pose as much risk as anthropogenic metals.
                                           18
                                                                                                _

-------
9.  The Framework discussion of environmental chemistry should emphasize the involvement of
metals in biogeochemical cycles.

10. The environmental chemistry section of the Framework should include a discussion of
processes affecting metals in sediments, and reactions that incorporate metals in organic
compounds such as methylation.

11. The "bioavailability" section of the Framework should be revised to define what is unique to
metals about bioaccumulation and how this information might be used in risk assessments.  EPA
should move the conceptual bioaccessibility/bioavailability model and related discussion to
Section 4 of the Framework.

12. The important Framework discussion of acclimation, adaptation, and tolerance should be
linked to the discussion of essentiality.

Long-term

14. EPA should revise the Framework to include a discussion of assessing the risks of
metal/metal contaminant mixtures as well as metal/organic contaminant mixtures.

6.2.2     Charge Question 2.2. Please comment on how well the conceptual model presents
         key metal processes and whether or not it is complete.

6.2.2.1   Comments in Response to Charge Question 2.2

   The SAB finds that the conceptual model in the Framework is sufficiently  comprehensive.
However, the conceptual model should be more clearly  linked to text in various parts of the
Framework. The SAB recommends revisions to improve presentation of the conceptual model
and to emphasize key concepts in the model.

Completeness of Conceptual Model

The conceptual model as depicted in Figure 2-3 of the Framework is sufficiently comprehensive.
It  is closely related to a conventional multimedia exposure model. A key difference between
metals and most organic compounds with respect to fate and transport is the biogeochemical
cycling of metals. The role of biogeochemical cycling in the conceptual model for metals does
not appear to be adequately represented in Figure 2-3, though it may be considered under the
"Environmental Chemistry" (Ml) part of the diagram.  At a minimum, the text related to Figure
2-3 should mention the role of biogeochemical cycling. As currently presented, the conceptual
model lacks the feedbacks involved in biogeochemical cycling.

Linkage of Conceptual Model to Text in the Framework

   The SAB notes that Figure 2-3 of the Framework is a compact summary of the  conceptual
model upon which the risk assessment framework is based. The text in the various parts of
Section 2 should therefore be related to Figure 2-3. This can be accomplished with some modest
                                          19

-------
revision of the existing text.  More detail will be needed in some parts in order to explain the
relevance of some of the components of Figure 2-3 not currently addressed in the text (e.g.,
transport models).  In revising the parts of Section 2 to explain linkage with the relevant
components of Figure 2-3, links to related parts of Sections 3 and 4 should be included where
appropriate.

  The SAB finds that Figure 2-2 of the Framework is also an important organizing graphic, but it
focuses on detailed processes that are not discussed in detail in Section 2 of the document. As
Section 2 is an overview of basic factors to be considered in metals risk assessment, Figure 2-2 is
too detailed to be included in this section.  Figure 2-3 provides the high level of aggregation
appropriate for Section 2. Figure 2-2 is well structured and informative, but should be moved to
Section 4 where it can be introduced and explained in detail, and linked to the topics discussed in
that part of the Framework document.

Key Concepts to be Emphasized in the Conceptual Model

   The conceptual model in the Framework is closely related to conventional organic multimedia
models, both in the component models chosen and  in the linear sequence in which they are
applied. Much of the Framework is devoted to distinguishing concepts used in metals risk
assessment from organic risk assessment.  The following key concepts that are not indicated in
the conceptual model diagram should be emphasized  either by modifying the diagram or by
adding accompanying text where Figure 2-3  is introduced:

    •   Precipitation/dissolution of mineral phases that contain a metal can lead to a decoupling
       of the usual linear relationship between the total mass of a metal in an environmental
       compartment and the free ion or other dissolved metal concentrations.

    •   Cyclical metal transformation processes, such as oxidation/reduction and
       methylation/demethylation, are not readily handled by organic fate and transport models
       since metal reactions do not result in  a permanent transformation to another compound.

    •   Natural loadings of metals differ from anthropogenic loadings in that they may come
       from inside the system of interest at rates controlled by natural processes.

    •   The fate and transport of both organic compounds and metais are coupled to the major
       biogeochemical cycles, such as carbon and nutrients.  In general, metals interact with the
       cycles of more elements (especially sulfur and other metals) than organic compounds.
       For organic compounds, the coupling to natural biogeochemicai cycles is essentially
       unidirectional (i.e., the rrajor biogeochemical  cycles affect the fate and transport of
       organics, but not vice versa).  For metals, exceptions to this rule are more common since
       metals can be limiting nutrients or toxicants to organisms that drive the major
       biogeochemical cycles such as higher plants, phytoplankton, or bacteria. This aspect of
       metal biogeochemistry cannot be simply accounted for in a linear framework. In  the
       absence of a comprehensive model, a means of allowing metals model outputs to  feed
       back into values selected for model input parameters that govern the major cycles may
       need to be devised.
                                           20

-------
   •   The "metalloregions" approach (briefly discussed on page 2-12 of the Framework) of
       defining "metal-related ecoregions" for regional or national-scale assessments is an
       evolving approach that may have merit.  Because no details on the approach are
       presented in the Framework, however, it is difficult for the reader to evaluate the strength
       of its potential value. The SAB recommends that an expanded description of the
       approach be provided, and that it be presented as just one example of how regional-scale
       risk assessment might be approached. The challenges that result from uncertainty and
       variability inherent in the approach should be addressed.

6.2.2.2   Key Recommendations in Response to Charge Question 2.2

Short-term

1.  Text related to the model depiction in Figure 2-3 of the Framework should mention the role of
biogeochemical cycling.

2.  Text in various parts of Section 2 of the Framework should be related to Figure 2-3. In
revising parts of Section 2 to explain linkage with components of Figure 2-3, links to related
parts of Sections  3 and 4 of the Framework should also be included.

3.  Figure 2-2 of the Framework should be moved to Section 4 where it can be introduced,
explained in detail and linked to the topics discussed in that part of the Framework.

4.  The following key concepts should be emphasized in the conceptual model by modifying
Figure 2-3 or adding accompanying text: precipitation/dissolution of mineral phases containing
metals, cyclical metal transformation processes, and natural loadings of metals.

Long-term

1.  Because metals can be limiting nutrients  or toxicants  to organisms that drive major
biogeochemical cycles, the conceptual model should incorporate feedback into model input
parameters that govern biogeochemical cycles.

8.  An expanded description of the "metalloregions" approach of defining "metal-related
ecoregions" should be incorporated into the Framework.

6.3.1     Charge Question 3.1.  Please comment  on how well the recommendations under
         Section 3 are supported by the detailed information in Section 4. Are there
         recommendations that should be included? Are there any inorganic metals or
         metal compounds for which any of the  recommendations would not apply?

 6.3.1.1  Comments in Response to  Charge Question 3.1

   The SAB has reviewed the recommendations in Section 3 of the Framework document and
provides  the following comments. To be most helpful, the recommendations should be tiered,
with the most critical general overarching recommendations (those with the greatest impact)
                                          21

-------
presented first, followed by other specific recommendations that would be of value to the
assessor. This would help focus the different sections of the Framework to ensure that the most
important issues are addressed. Tiered recommendations would also provide a platform for
subsequent documents so the assessor can prioritize how a risk assessment site is addressed.
Tiering options might involve restructuring all of Section 3 or developing tiered
recommendations within individual subsections. The SAB notes that recommendations
pertaining to various topics in the Framework are distributed throughout Section 3 of the
document.  For example, recommendations pertaining to environmental chemistry are included
in Sections 3.2.1 and 3.3.1 that present recommendations on environmental fate and transport.

General Comments on the Recommendations in the Framework

   The SAB provides the following general comments on the recommendations in the
Framework.

   •   To ensure that the document is not prescriptive, as stated in the document purpose in
       Section 1, the SAB recommends that prescriptive recommendations throughout the
       document be generalized. For example, instead of recommending a particular model or
       approach (such as recommendation 3  on page 3-24), the models should be described as
       alternatives among several approaches.

   *   Section 3 of the Framework should be reorganized to make it internally consistent with
       other parts of Section 3. For example, the headings for aquatic risk assessment should be
       more similar to those for terrestrial risk assessment. There is a lack of parallelism
       between the aquatic and terrestrial recommendations and balance needs to be achieved.
       The terrestrial recommendations, in general, include a broader range of approaches and
       include specific guidance to the risk assessor regarding the current state-of-the science
       (i.e., tools for today) as well as the direction of future tools and approaches.  A similar
       level of guidance and recommendations needs to be reflected in the aquatic discussion.

   •   Recommendations should be highlighted by minimizing textual justification, and cross-
       referencing the justification directly to those parts of Section 4 that support or discuss the
       recommendations.   Additionally, any  references to the scientific literature that are
       contained in the recommendations should be removed.  References should be provided in
       the sections of the Framework that support the recommendations.

   •   As opposed to the broad environmental chemistry recommendations given in Section
       3.2.1 of the Framework, the recommendations provided at the end of Section 3.3.1 (pages
       3-23 and 3-24) are very specific. The SAB notes that it is unclear whether this level of
       specificity is appropriate for a "Framework" document. A  greater degree of consistency
       is needed with respect to the specificity of the recommendations as a whole.

   •   In general, the environmental  chemistry recommendations  in Section 3 are supported by
       the discussion in Section 4.  However, it is difficult to determine which parts of Section 4
       correspond to particular recommendations in Section 3. In order to better assess the
       support for the recommendations in Section 3, it would be helpful to provide a "section
                                           22

-------
      identifier" indicating the source of the supporting information. Similarly, this might
      serve as a better way to organize Section 4.

   •  The focus of many of the environmental chemistry recommendations is on modeling.
      However, little information is provided on activities related to model validation or other
      data collection efforts that may be important for a given location. For example, the
      complex environmental conditions at a specific site may not be amenable to application
      of available models and may require substantial site-specific data.

   •  While it may be logical to separate the discussion of soil and sediment for the purposes of
      assessing exposure or toxicity, the SAB finds that from an environmental chemistry
      perspective it would be preferable to combine the discussion of the two media in one
      section.  In this format, geochemical origins and resulting similarities among soils,
      aquatic sediments, and subsurface sediments can first be highlighted. Risk assessment
      approaches that have evolved to depend upon  different factors such as controlling solid
      phases, solution composition, and redox conditions, can then be discussed.

   •  The SAB recommends that EPA reduce the number of specific recommendations in the
      Framework by omitting statements that are not recommendations and condensing similar
      or redundant recommendations.

Comments on Recommendations in Specific Sections of the Framework

  The SAB has numerous comments addressing the question of whether the recommendations
set forth in Section 3 of the Framework are directly supported by the more detailed discussion in
Section 4 and are justified or germane to an understanding of the risks of metals. The SAB
believes that some recommendations should be deleted and others amended. The SAB  also
believes that additional recommendations should be considered for inclusion in the Framework
document. Detailed comments, organized by section  number of the Framework, are provided in
Appendix A of this report.

6.3.1.2   Key Recommendations in Response to Charge Question 3.1

Short-term.

1.  EPA should provide tiered recommendations in the Framework. To be most helpful the most
critical recommendations should be presented first, followed by specific recommendations that
would be of value to the risk assessor.

2.  Prescriptive recommendations throughout the Framework should be generalized. Alternative
choices should be  described instead of recommending a specific approach or method.

3.  Section 3 of the Framework should be reorganized to make the parts of the Section  internally
consistent in scope and balance.

4.  The recommendations in the Framework should be highlighted by minimizing the amount of
text in Section 3, cross-referencing the justification directly to appropriate parts of Section 4 of
                                          23

-------
 the document.

 5. EPA should revise the Framework to provide a greater degree of consistency with respect to
 the specificity of the recommendations.

 6. With regard to recommendations concerning modeling, EPA should provide more
 information on model validation or data collection efforts that may be important for a given
 location.

 7. The discussion of soil and sediment should be combined into one section.

 8. EPA should reduce the number of specific recommendations in the Framework by omitting
 statements that are not recomrr endations and condensing similar or redundant recommendations.

 9. EPA should address detailed comments and suggestions provided in Appendix A of this
 report. The detailed comments in Appendix A focus on the question of whether
 recommendations in various sections of the Framework are directly supported by the discussion
 in Section 4 of the document.

 6.3.2    Charge Question 3.2. Please comment on the objectivity and utility of the data,
         tools, and methods discussed in Section 4.  Identify any scientific or technical
         inaccuracies, or any emerging areas or innovative applications of current
         knowledge that may have been overlooked or warrant a better discussion of
         uncertainty, including areas needing further research.

 6.3.2.1   Comments in Response to Charge Question 3.2

   The SAB finds that the huma:i exposure and health effects discussion in Section 4 of the
 Framework is not complete and contains errors. The SAB finds that this part of the Framework
 will require a major rewrite that may not be achievable in the short-term.  However, such a
 rewrite will be essential  if the treatment of human exposure and health effects is to be of equal
 value to other parts  of the Framework.  The environmental chemistry discussion in Section 4 of
 the Framework is comprehensive, but in many instances critical evaluations of the tools and
 methods are not provided and the justification for many recommendations is not clear. The
 ecological exposure and effects discussion in Section 4 of the Framework provides a great deal
 of supporting information for the recommendations articulated in Section 3 of the document.
 However, the treatment of various topics addressed in this part of the Framework is  uneven. The
 SAB provides recommendations for improvements in the document.

Human Exposure and Health Effects

   Much of the human health information in Section 4 was derived from the issue paper on
 human health effects of metals (EPA, 2004), which was not comprehensive and needs to be
expanded to improve this key resource for the Framework document.  The following are
examples of key items that need to be addressed.

   •   The SAB notes the importance of considering nanoparticles and their associated metal
       content in assessing human exposure to metals.  Dermal exposure  is also of considerable
                                          24

-------
      importance with regard to nanoparticles.

   •  The SAB notes that PMjo and PM2.s need greater attention as mixtures with regard to
      human exposure and health effects.

   •  The SAB notes that the discussion of Hg speciation was not given sufficient attention
      especially with regard to the source of exposure.  Additionally, Hg speciation in vivo is
      very complex and measurements of blood Hg levels generally do not distinguish
      between, for example, dental exposure to metallic Hg vapor and Methyl Hg from eating
      fish.

   •  There is reference in Section 4 of the Framework to the principle of metal accumulation
      in organisms that can be eaten by humans. The SAB notes that and this general principle
      applies to many metals but not all metals.  In vivo speciation of some metals may occur.
      For example, plants, fish, and game that take up hexavalent Cr can convert it to the less
      toxic trivalent form. Thus, humans can safely consume most plants and animals exposed
      to hexavalent Cr.

   •  The SAB notes the omission of any discussion in Section 4 of toxic effects of metals at
      low doses. This is a crucial issue because a number of metals exhibit a biphasic dose
      response curve with distinct adverse effects at low doses and a different type of toxic
      response at higher concentrations. The SAB recommends the inclusion of a section in the
      Framework that describes low dose toxic responses to metals and their compounds. For
      example, it is now apparent that Pb exhibits a biphasic dose-response curve with a greater
      slope of toxicity versus blood Pb concentrations at low exposure levels.

   •  The SAB suggests that Section 4 should include an analysis of: 1) the extent to which the
      use of Benchmark Dose Modeling decreases uncertainty and improves the derivation of
      RfDs for metals compared to the use of no observed adverse effects levels (NOAELs),
      and 2) the importance of updating current RfDs using the Benchmark dose modeling
      approach .

   •  The SAB notes the importance of including more summary tables in the Framework to
      enhance the understanding of the complex information presented in section 4.

   •  The SAB notes an insufficient discussion of the interactions between metals and organic
      chemicals as it applies to the problem of mixtures. There needs to be more discussion in
      the Framework of how metals interact with organics and how this interaction can lead to
      potentiation or antagonism. The SAB also notes the importance of applying proper
      objective criteria to assessing these interactions, including correct statistical tests.
Environmental Chemistry

   The SAB finds that, with respect to environmental chemistry, the coverage of available tools
for risk assessment and methods for metals analyses is fairly comprehensive, with an emphasis
on tools and  methods unique to metals.  Detailed descriptions of tools and methods are not given
in the Framework, but adequate references are cited.  In many instances however, critical
                                          25

-------
 evaluations of the tools and methods are not provided and the justification for many
 recommendations is not clear.  Two examples are given below:

    •  In recommending analytical techniques to characterize rnetal speciation (page 3-23, lines
       6-13), no evaluations were presented in Section 4 to help distinguish between methods
       commonly available through contract laboratories and those that presently are only
       available through research universities and laboratories.

    *  In recommending computer modeling to predict metal speciation in soil solutions (page
       3-23, lines 15-18), the computer programs Windermere Humus Aqueous Model
       (WHAM) and Non-Ideal Competitive Absorption Model (NICA) are cited without any
       discussion in Section 4 of the strengths, weaknesses, and limitations in the modeling
       approaches.

 The SAB therefore recommends that more emphasis be placed on developing comparative
 assessments of available tools and methods, and on providing additional information to aid risk
 assessors in deciding when particular tools and methods are and are not appropriate.

   The SAB also recommend:; that the balance of coverage in Section 4 be reviewed. The
 following issues are cited:

    •  Modeling tools, and to a lesser extent, analytical methods are included in the Framework.
       Limited information however is provided on what should be considered in data collection
       efforts (e.g.,  such as the type of data to be collected, appropriate temporal and spatial
       time scales to be considered, and data quality requirements that are unique to metals
       evaluations).

 •      EPA should provide a more balanced discussion of approaches for measuring  solution
       speciation versus techniques for assessing solid phase speciation.  In Section 4, no
       mention is made of current methods to assess free metal ion concentrations in  the solution
       phase for some metals directly (e.g., through specific ion electrodes, voltametry, or
       standard  EPA methods) or for measuring solution speciation for some metalloids.

 •      Although it could be argued that several of the modeling tools presented apply equally
       well to marine and freshwater systems, specific issues for the marine environment (e.g.,
       background concentrations and ion strength corrections) are barely addressed in
       comparison to the specific issues for freshwater environments.

   The SAB also recommends that Section 4 of the Framework contain.additional consideration
and discussion of data requirements and model uncertainty. Issues that should be addressed
include: criteria for designing a sampling plan, data requirements for model calibration, suitable
techniques for estimating parameter values (and associated uncertainties) for simple and complex
models, and evaluation of model uncertainty in model simulation results that are specific to
metals.

   In addition, the SAB recommends that biogeochemical cycles be discussed in Section 4 of the
                                           26

-------
Framework. This discussion should include the effects of organic carbon (and possibly iron and
manganese) on the fate of metals in the environment, and the effects of metals on organic carbon
and other ecological cycles through nutritional limitations or through toxic response.

Ecological Exposure and Effects
   With respect to ecological exposure and effects, the SAB finds that Section 4 of the
Framework offers a great deal of supporting information for the recommendations that are
articulated in Section 3. However, the manner of treatment for the various parts of Section 4
relative to ecological metals risk assessment should be more parallel in format across the
pathways of exposure.  This is particularly evident in the uneven treatment of topics such as
aquatic sediment and bulk sediment chemistry in comparison to the treatment of soils. A critical
shortcoming of the treatment of ecological metals risk assessment is the lack of a discussion of
levels of uncertainty, both in the knowledge base as well as in metals risk assessment
implementation. A discussion of uncertainty should be more explicit and more uniformly
distributed throughout the presentation of the current state of knowledge. The following specific
shortcomings are noted.
   The Biotic Ligand Model (BLM) approach is highlighted in Section 4 of the Framework but
the reader is never provided with a clear definition of the concept. The concept is treated as
though the reader is already familiar with this approach. It is the opinion of the SAB that the
BLM concept should be clearly  defined.  Trophic transfer is discussed extensively in Section 4 of
the Framework but not with respect to the BLM.

   There is very little attention given  in Section 4 of the Framework to the importance of
parameter, model, and laboratory validation in the field. There is a great deal of emphasis in this
section on models as tools for metals  risk assessment as appropriate, but the section lacks a
discussion of field validation needs and the consequences of this deficiency in the current state of
knowledge. There is little discussion  of ecosystem assessment or habitat assessment, the
discussion focuses on biotic indicators only.

   Section 4 of the Framework contains a good discussion of dietary exposure and  trophic
transfer but the tools to deal with these processes are not comparably developed. For example,
tools such as dynamic modeling (i.e.,  biodynamic or biokinetic modeling) should be included in
the discussion.

   Section 4 of the Framework should also include a discussion of acclimation and adaptation.
This is discussed in more detail  in the response to charge question 3.11.  Species sensitivity
distributions (SSDs), mentioned in  Section 4 but not discussed, should be more fully described in
this section. There is limited discussion of vascular plant risk assessment for metals and this
topic could be further developed in a  manner parallel to other sections of the chapter.

6.3.2.2  Key Recommendations in  Response to Charge Question 3.2

Short-term

 1. The discussion of mercury speciation in the Framework should be given greater attention,
particularly with regard to the source  of exposure.
                                            27

-------
2.  EPA should clarify the discussion of the principle of metal accumulation in organisms to
indicate that it does not apply 1.0 all metals; in vivo speciation changes must be considered.

3. A discussion of toxic effects of metals and metal compounds at low doses should be
incorporated into Section 4 of the Framework.

4. An analysis of the extent to which use of Benchmark Dose Modeling decreases uncertainty
and improves derivation of RfDs for metals compared to the use of no observed adverse effects
levels should be incorporated into the Framework.

5. EPA should include more summary tables in the Framework in order to enhance the
understanding of the complex information presented in Section 4.

6. EPA should place more emphasis in Section 4 of the Framework on comparative assessments
of a available tools and methods, and on providing additional information to assist risk assessors
in deciding when particular tools and methods are, and are not, appropriate.

7. In Section 4 of the Framework EPA should provide a more balanced discussion of approaches
for measuring solution speciation versus techniques for assessing solid phase speciation.

8. A discussion of biogeochemical cycles should be incorporated into Section 4 of the
Framework.

9. In Section 4 of the Framework EPA should provide a more balanced discussion of exposure
pathways relative to ecological risk assessment. The treatment of topics such as aquatic
sediment and bulk sediment chemistry is particularly uneven in comparison to the treatment of
soils.

10. The Biotic Ligand Model concept should be more clearly defined in Section 4 of the
Framework. Trophic transfer is not discussed with respect to the Biotic Ligand Model.

11. Section 4 of the Framework should contain a discussion of tools such as dynamic modeling
(i.e., biodynamic or biokinetic modeling)  to deal with dietary exposure and trophic transfer.

12. The ecological exposure and effects part of Section 4 of the Framework should contain a
discussion of acclimation and adaptation.

Long-term

13. The Framework should address the importance of considering nanoparticles and their
associated metal content in assessing human exposure to metals.

14. In the Framework, PMio and PMi 5 mixtures should receive greater attention with regard to
human exposure and health effects.
                                           28

-------
15. The Framework should contain more discussion of how metals interact with organic
compounds, and how this interaction can lead to potentiation or antagonism.

16. The Framework should contain more information on factors to be considered in data
collection for metals evaluations (e.g., type of data, temporal and spatial time scales, and data
quality requirements unique to metals).

17. The Framework should contain additional information on modeling issues to be considered
in the marine environment (e.g., background concentrations and ion strength corrections).

18. Section 4 of the Framework should contain additional information to address model
uncertainty and data requirements.

19. The Framework should contain a more explicit discussion of uncertainty in ecological risk
assessment of metals.

20. The Framework should contain a discussion of field validation needs for models.

6.3.3     Charge Question 3.3. Please comment on the state of the science (i.e., data, tools
         and methods) to address inorganic metals speciation in all environmental
         compartments for any given inorganic metal from the point of environmental
         release to the point of toxic activity as discussed in the document. Please comment
         on whether the framework identifies appropriate research needs to overcome any
         limitations in the state of the science.  Please address these questions separately
         for each of the three types of assessments presented (i.e., site-specific, national
         level, and ranking and categorization.)

6.3.3.1   Comments in Response to Charge Question 3.3

   The SAB finds that the discussion of speciation in the Framework could be improved by
bringing this information together and placing it in one part of the document. As discussed
below, the SAB also finds that additional information relevant to the subject of metals speciation
should be included in the Framework. In addition, the SAB finds that the research needs section
of the Framework appears to be a collection of limitations with no systematic or comprehensive
development of the limitations. The SAB therefore recommends that the limitations be discussed
in Section 4 of the Framework and that Section 5 of the Framework be removed from the
document.

   The SAB notes that  the major limitation in addressing inorganic metal speciation in risk
assessment is the lack of good  analytical tools for the direct measurement of metal
species/fractions. However, tools to directly measure metal species do exist and are improving
with time. The need to develop these tools, and the data to support modeling of speciation,
should be discussed in  the Framework.  The SAB feels that the Framework should not
recommend specific analytical tools, but it should discuss the importance of determining
speciation in environmental media and human biomonitoring samples.  The paucity of data to
support modeling of speciation limits the risk assessor's ability to adequately include speciation
in metal risk assessment tasks at site and national  scales. The SAB notes, however, that metal
                                          29

-------
speciation determination is more applicable for site-specific investigations than the setting of
national standards.

   The lack of analytical tools for direct measurement of metal species/fractions affects models
related to environmental transport and fate as well as exposure.  Section 5 of the Framework lists
research needs that would address this limitation but these needs are only listed in a bulleted
form. In comparison, the discussion of the Unit World Model, which may address other risk
assessment needs, is more extensive. The SAB notes that all research needs should be addressed
at a similar level of detail if this section of the document is to have relevance.  As it stands now,
this section is just a collection of limitations with no systematic or comprehensive development
of them. It would be preferable to include these limitations within the discussions of Section 4
and omit Section 5 of the Framework.

   The SAB provides the following specific comments in response to charge question 3.3.

   •   The SAB notes that it would be useful to collect the discussions of metal speciation in
       one location in the document.
   •   The Framework should contain a discussion of how to bound uncertainty in site and
       national efforts employing speciation.

   •   The SAB notes that a section needs to be added to the Framework on the importance of
       speciation of metals in human toxicity from the point of view of exposure and the
       diversity of species that can be formed within the body, (i.e., Cr (VI) and Cr (III), As
       methylation, elemental Hg and inorganic Hg, Cd metallothionein and other Cd ligands,
       etc.). It is important to identify the chemically and lexicologically active species of the
       metal as well.
   •   The SAB notes the importance of developing techniques to measure, in biological tissues,
       different species of metals to which humans can become exposed as well as to understand
       the species formed within the human body (e.g., methylated forms of As, and Cr
       oxidation states). This should be addressed  in the Framework.

   •   The SAB notes the importance of considering metal speciation for each individual metal
       since this concept makes sense only when considering each individual  metal.

   •   The SAB notes that numerous tools in the form of models and operationally defined
       analytical methods to address inorganic metal speciation are listed and discussed in the
       Framework. There are several well developed models for establishing the theoretical
       distribution of metals among species for given conditions in solution, although only
       specific ones are considered in much detail in the Framework. Similar models for
       understanding speciation in other media, such as soils and sediments, are not as well
       developed. As noted above, however, analytical tools to measure inorganic metal species
       are not very advanced.  Analytical tools that are discussed in the Framework (e.g.
       simultaneously extracted metals [SEM], sequential extractions) are, in reality, methods
       designed to fractionate ,in environmental matrix. With regard to application of these
                                           30

-------
      tools to the three types of assessments discussed in the Framework, models using the
      Hard and Soft Acids and Bases (HSAB) concept are probably most suitable for national
      assessments. The other tools appear to be applicable across the assessment types.

   •  The SAB notes that all discussions in the Framework that are related to speciation should
      adhere to the definition in the glossary. The use of consistent terminology when
      discussing forms of metals in various environmental matrices is recommended. This is
      discussed more fully in Appendix B of this report where a terminology proposed by an
      international expert body is provided.

   •  The SAB finds that the discussion of inorganic metals speciation is well developed in the
      Framework, and is successful in describing the importance of inorganic metals speciation
      in determining biological or ecological risk. However, the focus of the discussion is
      largely on the metal cations of greatest commercial interest, which represent only about
      one third of the metals of interest identified in the Framework scope (Section 1.2). The
      speciation section of the Framework should discuss all of the metals of interest,
      particularly the anionic metals Se, Sb, As, and V for which speciation is critically
      important in mobility  and toxicity. The discussion of inorganic metals speciation should
      also clearly address metals that do not behave like the metal cations.

   •  The discussion of speciation in the Framework should include a biogeochemical context
      which provides a more complete understanding of processes influencing metal exposure
      and metal transformations.  The discussion should point out where methods are available
      to directly measure metal species of interest and where modeling is the most suitable
      approach.

   •  The SAB finds that the Framework is selective in its treatment of speciation and
      transformations in the water column and in sediment, and would benefit from a more
      parallel organization of the discussion.

6.3.3.2   Key Recommendations  in Response to Charge Question 3.3

Short-term

].  The limitations in the research needs section of the Framework should be included in the
discussion in Section 4 of the Framework and Section 5 of the Framework should be removed
from the document. This revision is needed because the research needs section of the
Framework appears to be a collection of limitations with no systematic or comprehensive
development of the limitations.

2.  EPA should collect the discussions of metal speciation in one  location in the Framework.

3.  A section on the importance of speciation of metals in human  toxicity should be added to the
Framework.

4.  EPA should use consistent terminology when discussing forms of metals in various
environmental matrices.  The discussion in the Framework that is related to speciation should
                                           31

-------
 adhere to the definitions in the glossary.

 5. The Framework discussion of inorganic metals speciation should include all metals of interest
 (not just metal cations of greatest commercial interest), particularly the anionic metals, Se, Sb,
 As, and V, for which speciation is critically important in mobility and toxicity.

 6. The Framework discussion of speciation should include a biogeochemical context that
 provides a more complete understanding of processes influencing metal exposure and metal
 transformations.

 Long-term

 1. The Framework should contain a discussion of analytical tools for the direct measurement of
 metals species/fractions. The Pramework should not recommend specific analytical tools, but it
 should discuss the importance of determining speciation in environmental media and human
 biomonitoring samples.

 8. The Framework should contain a discussion of how to bound uncertainty in site and national
 efforts employing speciation.

 6.3.4     Charge Question 3.4. In an earlier draft of the framework, EPA had included
         three Summary Recommendation Tables in Section 3 on human health, aquatic,
         and terrestrial risk assessment, covering the three general assessment categories
         (i.e., site-specific, national level, and ranking and categorization). An example of
         this table is included as Appendix A in the draft provided to the SAB. To
         minimize confusion for users of the framework, the initial idea behind the
         recommendations and adjoining table was to have concise recommendations on
         the science, followed by a separate accounting of how these recommendations
         could then be applied to the different assessment categories. Reviews have been
         mixed on the utility of these tables as a sufficient communication tool. Please
         comment on whether tables of this type are useful for the final version of the
         framework. Does the panel have alternative suggestions for effectively
         communicating how the recommendations can be considered for each of the three
         assessment levels?

6.3.4.1   Comments in Response to Charge Question 3.4

   In general, the SAB finds that Summary Recommendation Tables such as the example
presented in  Table A-l of the Framework are a good way to summarize important points and
capture the structure character of the document. Tables  have an advantage as a way of
presenting a  summary, arranging complicated material to allow it to be viewed from different
perspectives, and facilitating organizing and cross referencing of materials.  However creation of
a summary table for a complex document such as the Framework is not straightforward. Issues
that arise include: the difficulty of representing complex concepts in short statements in the table;
the temptation to accept the abbreviated representation of the material in the table and ignore the
full complexity of the matter; and the fact that as the length and completeness of the table
                                          32

-------
increases, it expands across multiple pages and loses the advantage of a compact representation
of the material.

   The SAB recommends that the tables in the Framework be formatted differently and moved
forward to a "lead" position near the beginning of Section 3. The tables should be structured to
capture the recommendations presented in Section 3 in an organized manner that relates them to
their utility for the categories of risk assessment discussed in this document (i.e. national ranking
and categorization, national level assessments, and site specific assessments), with recognition
that within these three categories there are both screening and definitive risk assessments.  In this
regard, the tables should include key but limited information on currently available tools as well
as future tools and methods not yet readily available for use in operational risk assessment
activities. The tables should provide, at a glance,  an outline of the framework, key elements of
the framework recommendations, and available approaches (now and in the near future) to
accomplish these metals risk assessment goals. Use of the term "tools" as presented in the Table
A-l needs to be reassessed, because "fate and transport" and "bioaccumulation" are not specific
tools, but are aspects of risk assessment that require application of specific tools (e.g., extraction
techniques for estimating bioavailability).

   The SAB recommends that the tables not include references to the scientific literature but
 rather references to the specific parts of Section 4 of the Framework to explain the information
 and recommendations in the table. Table footnotes could be added to reference the relevant
 sections of the text and provide justification for each recommendation listed and summarized in
 the table. In this way, the tables become an operational, rapid index to the document. Such
 tables, if created for each of three broad subject areas covered in the Framework (i.e., ecological
 exposure and  effects, human exposure and health effects, and environmental chemistry), would
 help ensure consistency between the three areas.  These tables could be placed at the end of each
 relevant part of Section 3. The SAB notes that an alternative to using summary tables as a way
 of complementing the text would be to include a series of examples.

   Section 4 of the Framework should provide: 1) a thorough discussion of the background
science that supports the rationale for the framework structure and recommendations, 2) a
practical overview of current practice and the technical and political context of those activities,
and 3) a discussion of opportunities for improved approaches in metals risk assessment now and
in the future.  Section 4 should embody a state of science analysis that leads to sound assessment
practices, and  thereby highlights some of the logical research needs in the metals risk assessment
arena. Section 3 of the Framework should ideally include much less text, focusing on providing
the broader intent and context for the table.

   Table 1 below illustrates a possible approach to capturing the elements of the Framework into
a table, and providing a gateway to the information contained in Section 4 of the document. The
SAB notes that there are other approaches. It is suggested that the challenges in developing
tables will be  identifying brief descriptors of key elements of the Framework recommendations,
providing appropriate references to the sections of the document that fully discuss these
sometimes complex issues, and ranking otherwise complex and subjective aspects of information
related to each recommendation such as uncertainty. The SAB suggests that the overriding
benefit of developing  tables will be providing a visual summary of the essence of the Framework

-------
that offers information of value to metal risk assessors and directions to relevant information in
the document. Table 1 below is a skeletal representation of a possible table structure,-with an
ecological exposure and effects example filled in for the purpose of illustration.

6.3.4.2   Key Recommendations in Response to Charge Question 3.4

Short-term

1. Summary Recommendation Tables in Section 3 of the Framework should be reformatted and
moved to a lead position near the beginning of Section 3.

2. The summary recommendation tables should not include references to the scientific literature,
but rather references to the specific parts of Section 4 of the Framework that explain the
information and recommendations in the tables.

                                        Table 1.
DRAFT RECOMMENDATIONS TABLE - FRAMEWORK FOR METALS ASSESSMENT
No.

1
Ref
2
Ref

3
Ref
4
Ref
5
Ref

6
Ref
7
Ref
Recommendation





National
Screening
National
Ranking










BLM for use in
assessing
bioavailability
4.3.2




1





1
4.4.4.1




National
Complex





Site Scale
Screening




,
Site
Scale
Complex





Tools
Current





Tools
Future
Uncertainty
Data
Issues






1
4.8.1




1
X.X.X.X




1
x.x.x




1
X.X.X. X.X





X.X.X.X













1
XX




1
x.x.x






















.






















                                           34
                                                                                                      I

-------
Notes on Table 1:
6.3.5
Each recommendation in the table would occupy a block, but not all cells would be filled in. In most cases
Current and Future tools available would not be filled in, particularly when the recommendations deal
specifically with a tool. Where specific tasks are recommended, it is possible current and future tools
would exist.

Most cells would be filled by a numeric system where 0 = not good or not applicable, 1 = somewhat
available/applicable or other qualifier, and 2 = excellent option or application.  No verbiage is included in
the BLM example although it is possible that a few words might accompany the numbers in the boxes.

With each recommendation there is a row that includes references to the parts of Section 4 of the
Framework. Only single sections are listed although multiple pointers could be included in any box, and
should represent the roadmap to the relevant discussions.

Uncertainty is an important column even though it is highly subjective. Including this column highlights
the need to consider the uncertainty in the approach or tools being used by the risk assessor, and offers a
judgment of how much uncertainty is associated with that approach or factor or tool in risk assessment
could be due to natural factors, characteristics of the methodology, or other factors.

The data columns in the table can represent a  range of possible issues associated with data in risk
assessment, most often related to either the availability of the necessary data of the target metal for
different types of national or regional risk assessments, or the availability of ancillary data at any  scale that
is necessary to appropriately determine risk.

  Charge Question 3.5. Please comment on the objectivity of the Hard Soft Acid
  Base concept to applications of stability of metal complexes in toxicity
  assessments. See Section 4.1.2. (Emphasis added by SAB.)
6.3.5.1   Comments in Response to Charge Question 3.5

       Although charge question 3.5 specifically seeks comments on the objectivity of the Hard
Soft Acid Base (HSAB) concept to applications of stability of metal complexes in toxicity
assessments, the SAB  finds that the question could also apply more generally to risk assessment.
The SAB has commented on the objectivity of HSAB regarding both toxicity assessments and
the broader issue of risk assessment:

   The SAB concludes that the application of the Hard Soft Acid Base (HSAB) concept to the
stability of metal complexes in the context of risk assessment is generally presented in an
unbiased manner, with perhaps one possible exception. General statements that hard acids are
more toxic than soft acids should be worded more carefully to ensure that the statements are not
interpreted in a broader context than warranted by the available data.

    The application of the HSAB concept specifically to toxicity assessment is a more complex
issue. Whereas the HSAB concept is generally useful for assessing the strength of binding of a
metal to a receptor (if the chemical structure of the receptor is known), the extent of the toxic
response once the metal is bound is not really addressed by the HSAB  concept.  Clarification of
this distinction would  improve the objectivity of this section of the Framework.
                                               35

-------
   The clarity and completeness of the presentation could be improved by expanding the
 introduction with the following context for the application of the HSAB concept. The HSAB
 concept is a valuable way to summarize a considerable amount of qualitative chemical
 information and to allow the user to develop an intuitive feel for which complexes are likely to
 be more and less stable. The concept is well established in mainstream chemistry. However, the
 user should be aware that, while HSAB is useful for qualitative assessments of complex stability,
 quantitative calculations still depend on thermodynamic data such as stability constants and
 solubility products. These thermodynamic data are the basis of the models of metal speciation.
 The SAB also notes that additional citations to applications of the HSAB concept in
 environmental science would be useful (e.g., Sposito. 1989). In addition, the SAB recommends
 that, to ensure the accuracy of the presentation, the solubility constants in the Framework should
 be checked against established compilations of data.

 6.3.5.2  Key Recommendations in Response to Charge Question 3.5

 Short-term

 1. General statements indicating that hard acids are more toxic than soft acids should be worded
 more carefully to ensure that such statements are not interpreted in a broader context than
 warranted by the available data.

 2.  The Framework should indicate that the HSAB concept does not address the extent of the
 toxic response of a metal once it is bound to a receptor. The HSAB concept is generally useful
 for assessing the strength of binding of a metal to a receptor. However, the extent of the toxic
 response once the metal is bound is not really addressed by the HSAB concept.

 3. The introduction to the HSAB discussion should be expanded to  make users aware that while
 HASB is useful for qualitative assessments of complex stability, quantitative calculations still
 depend on thermodynamic data.

 4. Additional citations to applications  of the HSAB concept in environmental science should be
 included in the Framework.

 5. The solubility constants in the Framework should  be checked against established compilations
 of data.

 6. Specific revisions provided in Appendix C of this  report should be incorporated into the
 Framework to improve the clarity of the HSAB discussion.

6.3.6     Charge Question 3.6.  Please comment on the objectivity of the atmospheric metal
         chemistry discussion and its application to exposure assessments.  See Sections
         3.3.1.1 and 4.1.7. (Emphasis added by SAB.)

6.3.6.1   Comments in Response to Charge Question 3.6

   In responding to this charge question, the SAB notes that none of the Metals Risk Assessment
                                          36

-------
Framework Review Panel members has an active research program in atmospheric chemistry.
The SAB therefore recommends that an atmospheric chemist review these sections of the
Framework to ensure that there are no gaps in coverage, beyond those cited below.

   The SAB notes that there is no recommendation in Section 3.3.1.1 specifically addressing
exposure assessment. The one bulleted recommendation in Section 3.3.1.1 addresses models for
metal speciation in the atmosphere, and there is no text in Section 4.1.7 of the Framework to
support that one recommendation. The rest of the text in Section 3.3.1.1 is a summary of some
of the key points of Section 4.1.7, but it is not cast in the form of a recommendation.

   Section  4.1.7 of the Framework describes metals adsorbed to particles as the principle route
of direct exposure to metals in the atmosphere and cites the importance of particle size in
transport and exposure. The SAB concurs, within the limits of our knowledge of the subject, that
this assessment of direct exposure is generally accurate for most metals. However, the
discussion of atmospheric chemistry and its application to exposure assessment would be more
complete if the following issues were addressed.

•      A statement should be included about the potential for longer-scale transport of metals
       from a source through the atmosphere to soil, water, or air, from which exposure
       ultimately occurs. Even if the process for metals follows principles already established
       and described for organic compounds, and EPA does not want to repeat that description
       in the Framework, a statement about the similarities and differences between inorganic
       and organic compounds would improve the completeness of the Framework. For
       example, while many metals are transported in the atmosphere primarily only on the
       surfaces of particles, many organic compounds are transported in the atmosphere
       primarily as a component of the vapor phase.

•      A statement about the potential importance of volatile inorganic species of metalloids
      (e.g., H2S(g), AsH3(g)) should be included in the atmospheric chemistry discussion.

•      A statement about the potential importance of atmospheric transport to "background"
       concentrations of metals in the  environment should be included  either in the section of
       the Framework discussing atmospheric chemistry or in the "background" section.

6.3.6.2   Key Recommendations in Response to Charge Question 3.6

Short-term

\.  Because none of the Metals Risk Assessment Framework Review Panel members has an
active research program in atmospheric chemistry, an atmospheric chemist should review
Sections 3.3.1.1 and 4.1.7 of the Framework to ensure that there are no  gaps in coverage beyond
those identified below.

2.  Section 3.3.1.1  of the Framework should contain a recommendation  specifically addressing
exposure assessment.
                                          37

-------
3. The Framework should include a statement addressing the potential for long-scale transport of
metals from a source through the atmosphere to soil, water, or air from which exposure
ultimately occurs.

4. The atmospheric chemistry section of the Framework should contain a statement concerning
the potential importance of volatile inorganic species of metalloids (e.g., HbSCg), AsHs(g)).
5. The Framework should contain a statement concerning the potential importance of
atmospheric transport to "background" concentrations of metals in the environment.

6.3.7    Charge Question 3.7.  Please comment on the objectivity of the metal chemistry
         and environmental parameters incorporated in the various metal surface
         complexation and partition coefficient models and their applications to exposure
         assessments.  See Sections 3.3.1.2 and 4.1.4.1.

6.3.7.1   Comments in Response to Charge Question 3.7

  The SAB finds the Framework discussion of surface complexation and partition coefficient
models to be generally accurate and unbiased, but notes the following areas where the
presentation seems to lack completeness.

   •   The limitations of the models, particularly the data needs for the surface complexation
       models and the potential difficulty of obtaining the data, should be made more clearly
       obvious.  The SAB questions, for example, how realistic it is to propose routine
       application of surface complexation models in risk assessment

   •   A statement should be made in the Framework about the importance of balancing detail
       and uncertainty over ths entire assessment. The SAB questions, for example, whether it
       is appropriate to combine a detailed, molecular- level model of one process with an
       empirical, "black-box" model of another process, within the same risk assessment.

   •   A statement should be made in the Framework about the applicability of the surface
       complexation and partition coefficient models as a function of ionic strength, particularly
       with regard to estuarine and marine environments.

   •   The sediment chemistry and soil chemistry sections of the Framework should be
       coordinated to ensure that similar recommendations are given for similar circumstances.
       Combination of the text addressing environmental chemistry of soils and sediments into a
       single section should be: seriously considered (whereby it is recognized that ecotoxicity in
       the two environments should still be treated separately.)

   •   A statement should be made in the Framework to the effect that, if a K
-------
       evaluating the potential variability of K
-------
       baseline against which elevated levels from other natural or anthropogenic sources can be
       compared.

   The term background is often incorrectly assumed to connote "natural" and therefore "safe"
 or of no significant human or ecological health concern. However, ambient levels can vary, or
 can be inherently high enough to represent a potential health concern in and of themselves. They
 can also represent a total level from a combination of natural and anthropogenic sources, some of
 which may be historical or unlcnown. For metals in particular, the concept of background levels
 is complicated by several factors, as described in the Framework document, which include the
 sometimes highly variable natural levels of metals in soils, sediments, air and water, various
 historical anthropogenic sources or activities, and air deposition from distal anthropogenic
 sources.

   For example, natural levels of arsenic in soils can vary over a wide range from region to
 region depending on the sediment types from which the soils are derived, by as much as a factor
 of 10- to 20-fold. In addition, arsenical pesticides have been used over the past hundred years in
 agricultural and other settings; smelting and other air emissions can also contribute to local
 arsenic soil levels. Also, certain conditions, such as the chemistry of landfills, can lead to
 mobilization and release of natural sources of arsenic from rocks and soil, leading to greatly
 elevated arsenic levels in groundwater, but from entirely natural sources of arsenic. However,
 this can be distinguished from normal ambient levels with appropriate sampling and/or
 modeling. Use of the term "ambient" does not connote an ability to identify the various
 contributions from natural and anthropogenic sources, but does distinguish  between setting a
 benchmark level for a site or region against which other anthropogenic or anthropogenically-
 influenced inputs of concern can be measured. Anthropogenic metals can be those that are
 released into the environment from a specific human activity (i.e. a point source emission) or
 "natural" metals that may move from one environmental compartment  to another (i.e. soil to
 groundwater) due to a change in environmental chemistry related to a human activity.

   Since the concept of "background" is even more difficult to characterize in a human context,
 the SAB recommends defining and using the term "body burden" in this instance, since it is also
 a neutral term that attempts to quantify an individual's steady-state level using biomonitoring of
 one or more sample matrices (for example, blood, urine, hair, toenails,  bone scan, etc.). The
 Centers for Disease Control's (CDC) National Health and Nutrition Examination Surveys
 (NHANES) study is currently attempting to quantify and characterize body burdens in
 individuals so as to develop a national database that can serve  as the equivalent of a baseline
 measure against which the levels in an individual can be compared. Section 4 of the Framework
currently does not discuss this important issue. The SAB therefore recommends that the
 following definitions be added to the glossary and discussed in new sections in the human health
effects parts of Section 4.

Glossary recommendation:
   1.  Add - Body Burden:  An estimate of the concentration^) of a metal or metal species in
       specific tissues or the entire body, determined by the use of biological monitoring data in
       the appropriate matrix.
   2. Add - Human Biological Monitoring:  Use of measurements in specific  tissues or matrices
                                           40


-------
       (blood, urine, hair, toenails, bone, etc.) of specific metals or metal species in order to
       assess exposure or estimate body burden.

   The SAB also finds that Section 4 of the Framework does not adequately describe
biomonitoring. This is an important emerging area of public health evaluation and exposure
assessment that should be addressed. As with other aspects of metals analysis, speciation,
method of analysis, and choice of the appropriate matrix are critical aspects of effective
biomonitoring in humans. For example, analysis of chromium in blood, serum or urine does not
provide a way to distinguish between nutritional forms of chromium  from food or supplements
versus environmental or occupational exposures to hexavalent chromium that may be of concern.
Likewise, analysis of total arsenic in blood or urine does not reflect body burdens or recent
exposures to inorganic arsenic since food contains high but variable levels of organic arsenic
forms. However, arsenic in toenails provides both specificity for inorganic arsenic and an
integration of arsenic exposures and steady-state levels over several weeks or months of
exposure. Thus, metal-specific issues need to be considered for any biomonitoring program.
However, effective biomonitoring can provide excellent data on individual body burdens  that
may reflect both exposures of concern and potential health risks. The lack of discussion on this
topic is a serious deficiency of both Sections  3 and 4 of the Framework. The SAB strongly
recommends amending these sections to include this discussion, and  further recommends that the
EPA consider partnering with CDC through its ongoing NHANES and State pilot biomonitoring
programs in this important area.

6.3.8.2   Key Recommendations in Response to Charge Question 3.8

Short-term

1.  EPA should use the term "ambient" or "ambient levels" in the Framework rather than
"background" both in the glossary and throughout the text and recommendations. This is a strong
recommendation of the SAB. A recommended definition of "ambient levels" is provided in the
detailed comments above.

2.  The term "body burden" should be defined and added to  the Framework to describe the
concentration(s) of a metal or metal species in specific tissues or the entire body, determined by
the use of biological monitoring data in the appropriate matrix.

3.  The term "human biological monitoring" should be defined and used in the framework glossary
and text.  A recommended definition is provided in the detailed comments above. The fnsititute of
Medicine of the National Academies of Science has completed a substantial amount of work in this
area (National Academies of Science, 2004).

Long-term

4.  The SAB recommends including a discussion of the topic of biomonitoring in Sections 3 and 4
of the Framework, and also recommends that EPA consider partnering  in biomonitoring efforts
with the Centers for Disease Control through ongoing National Health and Nutrition Examination
Surveys (NHANF,S) and state pilot biomonitoring programs.
                                          41

-------
6.3.9     Charge Question 3.9. Please comment on the objectivity of the discussion of
         essentiality versus toxicity, including the relationship between Recommended
         Daily Intakes (RDAs) and thresholds such as Reference Doses (RfDs) and
         Reference Concentrations (RfCs).  See Sections 3.1,4.3.2, and 4.3.3

6.3.9.1   Comments in Response to Charge Question 3.9

   The SAB finds that a number of revisions are needed in the Framework to clarify and ensure
accuracy of the discussion of essentiality, Recommended Daily Intakes, and thresholds such as
Reference Doses and Reference Concentrations. The SAB provides the following comments and
recommendations in response to charge question 3.9.

   •   The SAB notes that for some metals, there might be an apparent discrepancy between the
       RDA and the calculated RfC or RfD. The EPA should consider the RDA for essential
       metals when considering the RfC/RfD. However, it should be noted that the RDA is
       usually satisfied by nonnal dietary intake of food, so that the RfC/RfD may be defined as
       a potential increment to the body burden of that metal from other dietary or extrinsic
       sources.

   •   The SAB notes a need to define  essentiality and, in this definition, to include the role of
       the metal in an essential physiological or biochemical process.

   •   The SAB notes that in Section 4.3.2 of the Framework it is important to restrict the
       discussion of essentiality to humans and to revise tables 2-1 and 4-12, which are
       identical. Table 2-1 could include a list of essential and non-essential metals in all
       organisms, with footnotes to denote those known to be essential in just plants, animals or
       humans. Table 4-12 should be restricted to a  list applicable solely to humans.

   •   The SAB notes that the current versions of Tables 2-1 and 4-12 need major revisions.
       The following recommendations apply specifically to the human table.  It is
       recommended that Mg be added to the list of nutritionally essential metals. In addition,
       the middle column of the table should be eliminated and the metals in that column moved
       to the third column that lists metals with no known beneficial effects. The metals in the
       second column that should be moved to the third column include: As, B, Ni, Si, V, Ba,
       and Sr. These particular metals should be noted  by asterisks in the third column to denote
       that there are limited human data for these metals.

   •   The SAB notes that a summary table should be added that includes RDA, RfDs, and
       RfCs available for the essential metals. The table should also include the adverse effects
       that occur at concentrations near or below the RDA for a given metal. This section
       should also specifically reference recent U.S.  Department of Agriculture (USDA) and
       National Research Council (NRC) reviews on essentiality of elements in humans.
                                          42

-------
6.3.9.2 Key Recommendations in Response to Charge Question 3.9

Short-term

1. The term, "essentiality" should be defined in the Framework.  The definition should address
the role of the metal in essential physiological or biochemical processes.

2. The discussion of essentiality in the Framework should be limited to humans. Table 2-1
could include a list of essential and non-essential metals in all organisms, and Table 4*12 could
include those applicable solely to humans.

3. The major the revisions of Tables 2-1 and 4-12 discussed in the detailed comments above
should be incorporated into the Framework.

Long-term

4. The Framework should be revised to consider the Recommended Daily Intake for essential
metals when considering Reference Doses and Reference Concentrations.

5. The Framework should contain a summary table providing Recommended Daily Intakes,
Reference Doses, and Reference Concentrations for essential metals. The table should also
include adverse effects that occur at concentrations near or below the Recommended Daily
Intake for a given metal.

6.3.10   Charge Question 3.10. Please comment on the objectivity of the discussion and
         recommendations presented for assessing toxicity of mixtures, including how to
         assess additivity versus departure from additivity (See sections 3.1.3.4 and 4.3.6).

6.3.10.1  Comments in Response to Charge Question 3.10

   The SAB finds that a number of revisions are needed in the discussion of toxicity of mixtures
that is contained in the Framework. The SAB provides the following specific comments and
recommendations in response to charge question 3.10.

   •   The SAB finds that the Framework discussion of the mixtures topic (Section 4.3.6) is
       limited and needs clarification and expansion. This section needs to be expanded to
       address co-exposures with organic pollutants (e.g., TCE, solvents, hydrocarbons) and air
       pollutants (e.g., gases such as ozone and particulates). The section needs more and
       improved examples of interactions for each of the conditions, and would benefit from a
       table that lists typical interactions and the ensuing effects on  toxicity.

   •   The SAB recommends that the example of the selenium and  mercury interactions on the
       bottom of page 4-78 be deleted. It is not an appropriate example since it leaves the
       impression that selenium supplementation should be used to  prevent mercury toxicity.

   •   The SAB recommends that the mixtures topics part of the Framework (currently Section
                                           43

-------
       4.3.6) contain subsections:
       a.) Exogenous non-essential metal(s) effect on nutritionally essential metals.
              i) effects via molecular/ionic mimicry
       b.) Interactions between non-essential metals
              i) effects via interactions at a common site
              ii.) effects via one metal affecting one site and another metal affecting another site
       c.) Interactions of metals with non-metals
              i.) interactions with organics
                     1) effects on toxicity of the metals
                     2) effects on toxicity of the organics
              ii.) interactions with gasses and/or particulates
                     1) affecting metal uptake
                     2) affecting metal toxicity

    •  The SAB  suggests the inclusion of a new Framework recommendation that states:
       "Metal mixture interactions and toxicity need to be clearly demonstrated by the use of:
       a.) proper experimental design (National Research Council, 1988)
       b.) appropriate plotting of diagrams
       c.) rigorous statistical evaluation to demonstrate synergy, additivity, potentiation, sub-
          additivity and/or antagonism."

    •  The recommendations in Section 3.1.3.4 of the Framework need to address the National
       Academy  of Sciences/National Research Council (NAS/NRC) Complex Mixtures report
       (National  Research Council, 1988). Recommendation 1 (page 3-9, line 9) in Section
       3.1.3.4 should address the NRC report.  Recommendation 4 (page 3-9, line 22) in Section
       3.1.3.4 should be rephrased to state: "There are established interactions that are based on
       metal mimicry. Future research goals should determine how considerations of metal
       mimicry affect risk assessments and metal toxicity."

    •  A definition of metal mimicry is needed in the glossary of the Framework.  The SAB
       suggests the following definition: "Metals that exhibit structural similarity which results
       in competition for essential receptors thus disrupting normal functions, such as chromate
       or arsenate substituting for sulfate or phosphate, lead replacing Ca or Zn, and Cd
       substituting for Zn or Ca."  It might also be helpful to include in Section 4 of the
       Framework a table that presents examples of well-established metal mimicry.   It is also
       important  to note that metals can profoundly  influence each other's biology through
       mechanisms other than mimicry.

6.3.10.2  Key Recommendations in Response to Charge Question 3.10

Short-term

1. Section 4.3.6 of the Framework should be expanded to address co-exposures with organic
pollutants (e.g., TCE, solvents, hydrocarbons) and air pollutants (e.g., gasses such as ozone and
particulates). More and improved examples of interactions for each of the conditions and  a table
listing typical interactions and effects on toxicity should be included in this section of the
                                            44

-------
Framework.

2. The example of selenium and mercury interactions on the bottom of page 4-78 of the
Framework should be deleted, because it leaves the impression that selenium supplementation
should be used to prevent mercury toxicity.

3. Additional sections (listed in the detailed comments above) should be included in the
mixtures topics part of the Framework.

4. A new recommendation should be included in the Framework stating that metal mixture
interactions and toxiciry need to be clearly demonstrated by the use of: proper experimental
design, appropriate plotting of diagrams, and rigorous statistical evaluation to demonstrate
synergy, additivity, potentiation, subadditivity, and/or antagonism.

5.  Section 3.1.3.4 should be revised to address the National Academy of Sciences/National
Research Council complex mixtures report (National Research Council, 1988).

6. A definition of metal mimicry (provided in the detailed comments above) should be included
in the glossary of the Framework. A table in Section 4 of the Framework should contain
examples of well-established metal mimicry.

6.3.11   Charge Question 3.11. Please comment on tbe objectivity of the discussion and
         recommendations concerning natural background, bioavailabiliry,
         bioaccumulation, biomagnification, and trophic transfer in both aquatic and
         terrestrial environments. See Sections 3.2.2 to 3.2.4,3.3.2, 4.4.3,4.5.4, and 4.5.6 to
         4.5.9.

6.3.11.1  Comments in Response to Charge Question 3.11

   The SAB finds that many aspects of the discussion in Sections 3 and 4 of the Framework are
objective and of reasonable utility for risk assessors. The level of detail seems appropriate for a
document of this type (i.e., screening level guidance document).  However, as discussed above,
there are parts of the document that could be  improved and there are issues of balance among
parts of the document that should be addressed.  For example, the discussions of
bioaccumulation, biomagnification and trophic transfer are confusing at times. Some of the
recommendations in Section 3 are inconsistent with the discussion in Section 4 and the issue
papers. The Framework brings up some very important issues reasonably well. But it also
seems to advocate some methods without reflecting important uncertainties, unknowns, or lack
of informed consensus in their base of scientific support. After revisions, the greatest utility of
the Framework will be its value as a statement of considerations unique to metals. The major
issues that should be addressed lie in: the need for balance in integrating sections, the imbalance
among recommendations, the need to integrate discussions of uncertainties, and some omissions.

Routes of Exposure

   The SAB notes that the discussion of dietary exposure and trophic transfer in Section 4 of the
                                           45

-------
Framework was not reflected in Section 3. Section 4.4.2.3 of the Framework discusses
limitations to the SEM-AVS approach that are not mentioned in Section 3 (see details below).
Both sections appear to be more conciliatory than analytical. The statement that "the most
widely used approach of assessing metal exposure in sediments is based upon EqP theory" is not
true. Many more agencies and scientists use the methods detailed in documents referenced by
National Oceanic and Atmospheric Administration (NOAA) and/or Canadian guidelines (Long
& Morgan, 1990; 1991; McDonald et al, 1996; McDonald et al., 2000). These methods and
concepts are discussed in the Chemistry section, but not mentioned in Section 4.4.2.3.

Natural Background

   In the Framework, "background" is defined as both natural and anthropogenic levels of metal.
This lack of clear definition confuses the issue.  As discussed in the response to charge question
3.8 above, natural background should be a consideration, but the Framework document treats it
as a non-issue.  In the Framework discussion of background, no medium (e.g., soil, sediment,
water) is specified but the issue of background is different in different media. This issue is
acknowledged to be complex when evaluating sediments if particle size is ignored and no
sediment cores are available. The Framework states that background concentrations can vary by
as much as five orders of magnitude. The SAB finds that five orders of magnitude variation in
metal concentration is most  likely an exaggeration when described in reports and literature.  In
part, large variances may be the result of using earlier sediment and water chemistry data, when
adequate" clean chemistry"  methods were not used. Mention of the EPA Storage and Retrieval
(STORET) database, with a caveat about quality assurance, does little to help the risk assessor.
STORET contains data that  could be incorrect by five orders of magnitude because it represents
earlier, non-clean chemical analyses. The Framework needs to emphasize the importance of
ultra-clean chemistry in determining all metal concentrations, but especially those values that
might be background.

   In the response to charge question 3.8 above, the SAB recommends that the EPA use the term
"ambient" or "ambient levels" rather than "background."  However, the Framework document
should distinguish between "natural" and higher-level anthropogenically-induced backgrounds.
In discussing ambient or background levels, the Framework needs to specify the need for
determining what background is and, consequently, what to consider. Using the term "natural"
likely complicates the task of defining a base concentration for comparison in metal risk
assessment. The Framework should provide guidance to establish an ambient or "background"
concentration that would be operationally defined for the assessment taking into consideration
realistic concentrations that often will reflect both natural and anthropogenic influences.
Acknowledging "background" concentrations becomes assessment-specific. For example, San
Francisco Bay sediments have high nickel concentrations stemming from historical times.
Arsenic at regional scales presents a similar situation.  Background concentrations ultimately
dictate the kinds of organisms, the nature of ecology, and types of chemistry at that site or
region.

Bioavailability

   Bioavailability is a useful concept and should be brought into the Framework
                                           46


-------
recommendations.  The Framework statement on the "bioavailable fraction" is very important,
both in terms of the science and moving the Framework forward.  The Framework document
does handle the concept of bioavailability more extensively than other aspects.  It is clear that in
the view of the document "bioavailability" concerns speciation and other water chemistry
effects.  "Trophic transfer," "dietary exposure," and "biomagnification" are mentioned in a few
places, but there is little in  Section 3 of the Framework to help the risk assessor understand and
employ these concepts. The discussion of dietary toxicity leaves out important examples and
understates the importance of this route of exposure, as well as the increasing knowledge of it.
As the SAB has previously noted (U.S. EPA SAB, 2000a), dietary exposure is an important
consideration, or at least uncertainty, in any assessment of the ecological risks of metals. The
Framework does not adequately integrate this uncertainty into the overall view it presents to risk
assessors. There is no integrated view of how an organism might respond to all sources in
different circumstances; the routes of exposure are treated as if they are not related.

   A problem associated with the hazard assessment of metals in water is that very small
deviations from background concentrations result in very large amplification through the
environment because of high K<)S and relatively high BCF/BAFs for many metals. No guidance
concerning this issue is provided in the Framework document, nor is the essence of the issue
discussed at any length. Bioavailability as shown in the conceptual model  in the Framework
should include both exposure and dietary uptake. The Framework text provides an uneven
approach and should be expanded to address the influence of dietary uptake. The conceptual
model in Figure 2-2 of the  Framework includes dietary uptake, as it should, and provides a
rationale for including food type and food choice. However, there is also an ecological need to
incorporate dietary uptake  into the Framework discussion. There should be an emphasis in the
Framework on the need to  understand species presence and the nature of the food web. Trophic
transfer, for example, has been shown to be an important route of uptake of metals from
sediments into fish via planktonic invertebrates and into epibenthic invertebrates feeding on
peri phy ton.

Bioaccessibility

   The  SAB finds that "bioaccessibility"  is properly considered in the document and represents
the labile portion of the metal.

Bioaccumulation

   Bioaccumulation is a concept that is different from biomagnification. This presents some
level of confusion in  the discussion of the different levels of risk assessments in the Framework.
The important point that should be made  in the Framework is that metals bioaccumulate, and
trophic transfer is important. It is less important that biomagnification through the food  web is
likely to occur only in some circumstances (although examples  exist for selenium and
methylmercury).

   Bioaccumulation should be reviewed in the Framework as a concept for use in risk
assessments, particularly in the site-specific risk assessments. The issue of what construct to use
to express bioaccumulation (e.g., BCF, BAF, models) is separate from consideration of the
                                           47

-------
 bioaccumulation processes. Sections 3 and 4 of the Framework place great emphasis on the
 limits of a ratio approach and little emphasis on bioaccumulation processes that are relevant to
 exposure analysis in a risk assessment. A concern of the SAB is that coefficients in the ratios are
 not independent of exposure concentrations. The coefficients are calculated and used but they
 are highly variable. The concept of using BAF or BCF ratios can be appropriate, but it should
 never be assumed that they are constant(s), as is the typically assumed in uses like hazard
 assessment. This issue is further discussed in the response to charge question 3.12 below. The
 SAB recommends that a text box be included in the Framework document to highlight concept
 of BCF versus the use of this as a tool in site specific or national assessments. The SAB also
 finds that  there is a strong need for presentation of a conceptual model of bioaccumulation in the
 Framework. Such a conceptual model should tie bioaccumulation to toxicity.  If
 bioaccumulation and bioconcentration factors are treated more comprehensively, the Framework
 will be a more cohesive document.  The SAB's discomfort with the treatment of BCF and BAF
 has to do with difficulties in measuring bioaccumulation, which involves estimates of uptake,
 depuration, etc. Any method that can be related to a dynamic intake, and that relates site of
 target toxicity with effects, would be of value. Such models need to be better incorporated into
 the bioaccumulation discussion in the Framework.  Until this information is incorporated into the
 document, toxicity tests will be utilized, or concentrations in tissues will be used, without any
 understanding.

 Essentiality

   The SAB finds that the discussion of essentiality in the Framework also needs to be
 expanded, particularly with regard to how essentiality influences accumulation factors. Tissue
 concentrations can vary by a large amount and there is a need to discuss the factors influencing
 the site-specific characteristics  that  lead to a given BCF.  In this regard the Framework
 document should discuss the state of the science (versus what might simply be included in a
 Framework).

 Concentrations in  "Metallo-regions "

   The SAB recommends using a geometric progression (log-normal distribution) for metal
 concentrations in either "metal lo-regions" or catchment basins and  describing the low-end of the
 distribution (e.g., 95th or 90th percentile exceedence zones) as potential problem areas.  In a
 national or ranking risk assessment, a conservative approach would need to be taken by using
medians and the 90* (or so) percentile. For national-level risk assessments, one would
 necessarily want to err on the side of conservatism.

 Critical Body Residues

   The SAB finds that the concept of critical body residues (CBR)  is handled unevenly in the
Framework and is over emphasized. The fact that CBR can be measured does not necessarily
mean it is the concentration at the site of toxic action. Further, there are few data on this and it
has been measured in only a few species. The concept may be an idea that can be used in the
future.
                                           48

-------
Acclimation

   The SAB notes that there is much discussion in the Framework of acclimation and adaptation.
The costs of adaptation are discussed well in Section 4, but that discussion does not appear to
carry over to section 3. It is well known that organisms have developed a variety of
physiological and/or biochemical strategies for dealing with metals exposure due to the
ubiquitous presence of metals in the natural environment. In many cases these strategies have
permitted organisms to survive and thrive in areas where they would not normally be able to
exist. The importance of considering these strategies has long been debated among the
regulatory and regulated communities. It is also true that many metals are essential for the health
and development of organisms, and in some cases it has been observed that organisms used in
toxicity tests that have been cultured in "metals-deficient" media have been shown to be more
sensitive to subsequent metals exposure than are wild organisms raised in natural environmental
conditions. The general recommendation that has come from the scientific community is that
researchers should ensure that organisms used in conducting toxicity tests are cultured (or at least
acclimated for a period of time) to test media that contain metals concentrations that are
''similar" to natural background concentrations, not concentrations similar to the site in question.
It is assumed that this approach will reduce the potential of overestimating toxicity from "metals-
deficient" stressed organisms, while ensuring that underestimations of toxicity are not reached
from tests conducted with "metals-acclimated" organisms.  To this end,  it is equally important
that risk assessors are mindful of this potential concern and consider it in conducting their
evaluation of effects data.  The SAB finds that the discussion and recommendations contained in
Sections 3 and 4 relative to this issue do not adequately describe and delineate the difference
between true metals acclimation and test organism stress due to metals deficiency.

6.3.11.2 Key Recommendations in Response to Charge Question 3.11

Short-term

 \.  In addition to the discussion of EqP-based methods for assessing metal exposure in
sediments, Section 4.4.2.3 of the Framework should contain a discussion of other methods
(referenced above) for assessing metal exposure in sediments.

2.  The Framework should emphasize the importance of ultra clean chemistry in determining all
metal concentrations, but especially those values that might be  "ambient" levels. As discussed in
the response to charge question 3.8, the Framework does not provide a clear definition of
"background" levels of metals, and the SAB recommends using the term "ambient" or "ambient
levels" rather than "background."

3.  The Framework should distinguish between "natural " and higher-level anthropogenically-
induced ambient concentrations of metals, and provide guidance to establish an ambient or
"background" concentration that would be operationally defined for an assessment, taking into
consideration realistic concentrations that often will reflect both natural  and anthropogenic
influences.

4.  The concept of bioavailability should be brought into the Framework recommendations.  The
Framework should provide information to help risk assessors understand and  employ the
                                           49

-------
concepts of "trophic transfer, ''dietary exposure," and "biomagnification."

5.  The Framework text should be expanded to address the influence of dietary uptake of metals.
There is an ecological need to incorporate dietary uptake into the Framework discussion.

6. Bioaccumulation should be reviewed in the Framework as a concept for use in risk
assessment, particularly site-specific risk assessment. A text box should be included in the
Framework highlighting the BAF or BCF ratio concept versus its use as a tool in site specific or
national risk assessments.

7. The discussion of essentiality in the framework should be expanded, particularly with regard
to how essentiality influences accumulation factors.

8. EPA should use a geometric progression (log-normal distribution) for metal concentrations in
either "metallo-regions" or catchment basins and describing the low end of the distribution as
potential problem areas.

10. The Framework should be revised to state that, although critical body residues can be
measured, they do not necessarily reflect concentration at the site of toxic action. The
Framework should also indicate that critical body residues have only been measured in a few
species. The SAB notes that the concept of critical body residues is handled unevenly in the
Framework and is over emphas ized.

11. The discussion and recommendations in Sections 3 and 4 of the Framework concerning
acclimation and adaptation should be revised to describe and delineate the difference between
true metals acclimation in test organisms and test organism stress due to metals deficiency.

6.3.12    Charge Question 3,12. Please comment on the objectivity of the framework
         statement that the latest scientific data on bioaccumulation do not currently
         support the use of bioconcentration factor (BCF) and bioaccumulation factor
         (BAF) values as generic threshold criteria for hazard classification  of inorganic
         metals (see recommendation on page 3-17, lines 27-29 of the document). By this,
         the framework means that various assumptions underlying the BCF/BAF
         approach, including the independence of BCF/BAF with exposure concentration
         and the proportionality of hazard with increasing BCF/BAF do not hold true for
         the vast majority of inorganic metals assessed. Please comment on  the
         framework's acknowledgement that the appropriate use of BCFs/BAFs to
         evaluate metal bioaccumulation, including the degree to which BCFs/BAFs are
         dependent on exposure concentrations, needs to consider information on
         bioaccessibility, bioavailability, essentiality, acclimation/adaptation, regulation of
         metals (uptake and internal distribution), detoxification and storage, dependence
         on exposure concentration, and background accumulation.  While the ability to
         quantitatively address ail these factors may be limited at the present time, the
         framework states that their potential impacts should at least be qualitatively
         addressed. See Sections 3.2.4, 3.3.2.5, and 4.5.8.
                                          50

-------
6.3.12.1  Comments in Response to Charge Question 3.12

   The SAB agrees with the statement that BCF/BAFs do not apply for metals. The language of
the Framework is useful in describing the context of use for BCF/BAF. As stated in the
Framework, it is appropriate largely for use in a site assessment. The Framework acknowledges
that these methods may not be the best approach for use in a national assessment, and especially
for hazard rankings. However, guidance is offered in the Framework on how to derive
BCFs/BAFs (e.g., pages 3-17, and 3-33).  The SAB finds that the Framework document needs a
clearer discussion of when to use these tools, their deficiencies, and when they should not be
used. The justification of why or why not to use them needs to be more explicit and coherent.

BCF/BAF

   The SAB notes that the Framework does not mention that BCF/BAFs vary 50 fold or more
for every metal, partly because of inherent biological diversity in response to metals. A careful
analysis of the literature would show alternatives to the BCF/BAF approach that are much more
flexible and less variable (e.g., biodynamic models).

   The Framework correctly assesses the state of the science. Section 4.5.8 of the Framework
clearly expresses the issues and identifies shortcomings of the BCF/BAF approach.  There is a
difference in the utility of the BCF/BAF approach for assessing the risks associated with
organics and inorganics, and the Framework appropriately addresses these differences.  The SAB
supports the call for more data on bioavailability, acclimation, storage, metal regulation, and
accumulation as modifiers of BCF or BAF.  There  is no doubt that better data on metal  storage,
disposition in the body, and consequent potential toxicity, will result in better predictions of risk.
However, in some cases where data are limited, a precautionary stance of using potential BAFs
might be called for and not simply ignored.  For example, it would be  much clearer if the
Framework were to state that the BCF/BAF does not work for national assessments but it has
value for site-specific assessments.

   The Framework should specifically address the issue of hazard assessment and consider
trophic transfer.  The Framework needs to consider options beyond dissolved metals toxicity
tests. In this regard, the SAB suggests considering options that address: 1) the potential for
trophic transfer, and 2) the potential for transformation into bioavailable organometal
compounds.

6.3.12.2 Key Recommendations in Response to Charge Question 3.12

Short-term

 1.  The Framework should contain a clearer discussion of when to use BCF/BAFs, their
    deficiencies, and when they should not be used. The justification of why or why not to use
    them should be more explicit and coherent.

 2.  Assessment options beyond dissolved metals toxicity tests should  be discussed in the
    Framework.  In this regard, the SAB suggests that EPA consider options that address the
                                           51

-------
    potential for trophic transfer and the potential for transformation into bioavailable
    organometal compounds.

6.3.13   Charge Question 3.13.  Given the variety of organism responses to inorganic
         metals exposure, based on factors such as bioaccessibility, bioavailability,
         essentiality, uptake/excretion mechanisms, and internal storage/regulation, as
         described in Section 3.2.4, the framework states that BAFs/BCFs should be
         derived using mathematical relationships that represent the concentration in the
         organism or tissue as a function of the bioavailable concentration in the exposure
         medium/media for each set of exposure conditions. Please comment on whether
         this is the best approach based on the current state of the science or if there are
         alternative approaches that are more appropriate that can be routinely applied.
         See Sections 3.2.4,3.3.2.5, and 4.5.8.

6.3.13.1  Comments in Response to Charge Question 3.13

   The SAB finds that the mathematical relationships representing the metals concentration in
the organism or tissue as a function of the bioavailable concentration in the exposure
medium/media for each set of exposure conditions seem appropriate. However, the SAB
provides the following specific comments.

   •   Section 4.5.8 of the Framework indicates that steady-state conditions are often the
       primary concern in metals risk assessments, yet there certainly can be instances of non-
       steady state conditions being of primary concern (e.g. episodic hydrologic events and
       related metal mobilization). Further, if the recommendations to not apply BCFs and
       BAFs are supported, the SAB questions why recommendations to derive them are
       included in the Framework.

   •   The SAB finds that Sections 2 and 4.5.8.1 of the Framework clearly articulate issues
       surrounding the derivation and utility of BCF/BAFs for metals. For all of the reasons
       discussed in these sections, it appears that the concept of the BCF/BAF for metals holds
       little utility in assessing the environmental toxicity of metals in hazards rankings. One
       optimal approach (least uncertain) for deriving these values would be to use the tissue
       concentration at the site of action and to relate this to the best estimate of the biologically
       available metal. However, few data exist to allow derivation of such a value.

   •   The SAB notes that one aspect not mentioned in  Section 3.3.2.5 of the Framework is the
       use of multi-species model ecosystems to verify BAF or BCF predictions. Often the
       results of such real-world situations are to modify the growth (hence uptake and effects of
       metals).  The effective rate of uptake is very important, as the Framework states.  Hence,
       BAFs are not necessarily of value, as equilibrium situations are rarely found. The ideal is
       to have concentration measures at site of action and in the surrounding environment, but
       adequate tools are not immediately and widely available. Thus, the utility of the current
       construct is limited.  If the Framework were to include bioaccumulation dynamics, the
       variability would be narrowed.  On a site-specific basis, the ratios are better used than  in
       national assessments because variability may be less. For organics there are some well
                                          52

-------
      known and accepted assumptions.  For metals there is a large variability around the
      BCF/BAF estimates. However, there is little guidance as to "where to draw the line."

   •  The SAB strongly concurs that one cannot use a BAF or BCF ratio for national
      assessments or hazard ranking procedures. The SAB feels that a bioenergetics approach
      offers valuable potential for understanding metal accumulation from air, sediments, soils
      or water (Wang et al., 1996; Schlekat et al., 2002).  In the interim, the Framework should
      address metals bioaccumulation empirically for site assessments.  In the future, there
      should be a concerted attempt to generate data at the site of action (Escher et al., 2004).

6.3.13.2 Key Recommendations in Response to Charge Question 3.13

Long-term

1.  The SAB  strongly concurs that one cannot use a BAF or BCF ratio for national assessments
or hazard ranking procedures for metals and recommends that in the long-term EPA should
incorporate a bioenergetics approach into the framework.  Such an approach offers valuable
potential for  understanding metal accumulation for air, sediments, soil or water.

6.3.14   Charge Question 3.14. Please comment on the objectivity of the information and
        recommendations pertaining to the use of the acid-volatile sulfide-simultaneously
        extracted metals (AVS-SEM) approach and the biotic ligand (BLM) model. Are
        additional recommendations warranted? If yes, what are they? See Sections
        3.2.6,4.4.2.3, and 4.5.10.

6.3.14.1 Comments in Response to Charge Question 3.14

   It is the opinion of the SAB that the concepts of SEM-AVS and BLM are clearly on the
agenda for adaptation into risk assessment. The Framework comprehensively describes the
theory and evidence behind both methods.  However, the discussion in the Framework is
unbalanced throughout in comprehensively evaluating the practical and theoretical challenges
and inherent limitations that have been encountered in implementing the use of SEM-AVS
(Cantell, Burgess & Kester, 2002).  The primary  literature contains a number of questions that
are relevant with regard to implementation of SEM-AVS in either risk assessment or regulation.
The theory itself is attractive and a strong  literature supports its effectiveness in the environment
of the typical sediment bioassay. There is no question that sulfides are important in metal
associations  in anoxic sediments, or that sulfides control pore water metal concentrations in bulk
sediments. The questions about implementation  of the methodology lie in how the complex
vertical  gradients of sediments will be sampled, how stable SEM-AVS characterizations will be
for a site, and/or whether an SEM-AVS characteri/ation will hold for sediments that are moved
during normal  resuspension, flood or bed-load transport events. The SAB has provided similar
comments in a review of EPA's Integrated Approach to Metals Assessment in Surface Waters
and Sediments (EPA SAB, 2000a). Many such questions were also raised in a very  important
review of sulfide dynamics and its relationship to the stability of AVS in a cover article of
Environmental Science and Technology (Morse & Rickard, 2004).
                                          53

-------
   A second issue is that, although AVS controls bulk pore water concentrations, it does not
 control metal concentrations in what an animal eats. The literature that considers dietary
 bioaccumulation from sediments raises important issues with regard to the design of most
 sediment bioassay experiments: the living nature of sediments and how that affects
 bioavailability, and the biases that can occur in sediment bioassays of the type typically used for
 the AVS concept. These issues are not necessarily resolved one way or the other, but they are
 substantial and well enough documented that risk assessors must be made aware of the debate
 and be prepared to consider the pros and cons of the SEM-AVS method in a balanced way (Lee
 et. al., 1988; Lee & Luoma, 1998).

   The BLM is in the relatively early stages of development and also has inherent limits. For
 example, the BLM:  1) has no dietary component; 2) has no chronic component; and 3) has no
 cross-species comparisons among  differing mechanisms for binding and  effects-level meta!
 concentrations. Similar comments were provided in the SAB review of the Biotic Ligand Model
 of the Acute Toxicity of Metals (EPA SAB, 2000b) The published literature on animals other
 than trout and fathead minnow show simple, and not unexpected, correlations between toxicity
 test outcomes and metal speciation, in the guise of a biological model. The BLM definitely does
 account for speciation better than any methods to date; but the BLM does have limits at the
 present state of knowledge.

   The  information presented in the Framework regarding the use of the  BLM and SEM-AVS
 approaches is appropriate and reflects the current state-of-the-science. It is, however, interesting
 that the use of techniques relating to bulk sediment concentrations are conspicuously absent, at
 least in  terms of their applicability to large scale assessments. Methods such  as sediment quality
 criteria  (SQC), threshold effect level (TEL), and probable effect level (PEL) have a good role in
 conducting metals risk assessment, especially when data are not available to address metals
 sediment toxicity through methods such as SEM-AVS. Further, the implied lack of
 bioavailability of metals associated with sulfides has come into question  (Lee et a!., 2000). For
 risk assessments of a broader nature, e.g., at the national level, clearly the only viable approach
 to be implemented may be through the assessment of bulk sediment numbers.

   The SAB notes that the future of toxicity testing is moving toward mechanistic approaches
 and the  BLM approach  is a step in the right direction. An  important feature of the BLM  is that it
 addresses the site of action.  For chronic effects, BLM may not apply since site of effect  may
 change with exposure time frame.  The risk assessor has to be aware that there is not an available
 comprehensive tool and that there are limits to each approach. However, advances that move
 risk assessment toward consideration of bioavailable fraction, mode of action, and a mechanistic
 approach are steps in the right direction.

 6.3.14.2 Key Recommendations  in Response to Charge Question 3.14

Short-term

 1. The Framework should be revised to provide a more detailed discussion of the practical and
theoretical challenges and inherent limitations that have been  encountered in implementing the
 use of SEM-AVS.
                                           54

-------
2. The Framework should be revised to provide a more detailed discussion of the inherent limits
of the Biotic Ligand Model discussed above.

3. The Framework should present corresponding information on the practical challenges and
inherent limitations of using bulk sediment chemistry assessment methods.
                                           55

-------
 7.     REFERENCES

 Ahrland, S., J. Chart, and N.R. Davies.  1958. The relative affinities of Hgand atoms for acceptor
 molecules and ions. Quart. Rev. ChemSoc., 12:265-276.

 Cantell, M. G., R.M. Burgess, and D.R. Kester. 2002.  Release and Phase Partitioning of Metals
 from Anoxic Estuarine Sediments during Periods of Simulated Resuspension. Environ. Sci.
 Technol., 36(24): 5328-5334.

 Escher, B. I., and J.L.M. Hermans.  2004. Internal exposure: linking bioavailability to effects.
 Environ. Sci. Techno!.,  38(23): 455A-461A.

 Hathcock, J.N.  1996. Safety limits for nutrients.  J. Nutr. 126 :2386S-2389S.

 Lee, B-G., S.B. Griscom, J-S. Lee, H.L. Choi, C-H. Koh, S.N. Luoma, and N.S. Fisher. 2000.
 Influence of dietary uptake and reactive sulfides on metal bioavailability from sediments.
 Science, v. 14, 287: 282-284.

 Lee, B-G., and S.N. Luoma. 1998. Influence of microalgal biomass on absorption efficiency of
 Cd, Cr, and Zn by two bivalves from San Francisco Bay. Limnology and Oceanography. 43:
 1455-1466.

 Long, E.R. and L.G. Morgan. 1990. The potential for biological effects of sediment-sorbed
 contaminants tested in the National Status and Trends Program. NOS. OMA 52. Technical
 Memorandum. Seattle, Washington.

 Long, E.R., and L.G. Morgan. 1991.  The potential for biological effects of sediment-sorbed
 contaminants tested in the National Status and Trends Program. NOAA Tech. Memo. NOA
 OMA 52, Seattle, Washington.

 Lukaski, H.C.  1999. Chromium as a supplement. Annual Rev. Nutr.  19:279-302.
 Mac Donald, D.D., R.S. Carr, F D. Calder, E.R. Long, and C.G. Ingersoll. 1996. Development
and evaluation of sediment quality guidelines for Florida coastal waters.  Ecotoxicol. 5:253-278.

MacDonald, D.D., C.G. Ingersoll, and T.A. Berger. 2000.  Development and evaluation of
consensus-based sediment quality guidelines for freshwater ecosystems.  Arch. Environ.
Contam. Toxicol. 39:20-31.

Mertz, W. 1993. Chromium in human nutrition: a review.  J. Nutr.  123:626-633.

Mertz, W. 1995. Risk assessment of essential trace elements: new approaches to setting
recommended dietary allowances and safety limits. Nutr. Rev. 53:179-185.

Morse, J.W. and D. Rickard. 2004. The influence of sedimentary acid volatile sulfide (AVS)
chemical dynamics on toxic metal bioavailability.  Environmental Science &  Technology, 38,
 131A-136A.
                                          56

-------
National Academies of Science. 2004. Environmental Health Indicators, Bridging the Chasm of
Public Health and the Environment - Workshop Summary. National Academies of Science,
Institute of Medicine, Roundtable on Environmental Health Sciences, Research, and Medicine,
Board on Health Sciences Policy. National Academies Press, Washington, D.C.
httD://www.naD.edu/books/03090926S5/html/Rl.html.

National Academies of Science. 2000. Dietary Reference Intakes for Vitamin A, Vitamin K,
Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel,  Silicon,
Vanadium, and Zinc. National Academies of Science, Institute of Medicine, Food and Nutrition
Board, Panel on Micronutrients and Standing Committee of the Scientific Evaluation of Dietary
Reference Intakes. National Academies Press, Washington, D.C.
http://books.nap.edu/catalog/10026.html

National Research Council.  1988. Complex Mixtures: Methods for In Vivo Toxicity Testing.
National Academies Press, Washington, D.C.

Pearson, R G. 1963. Hard and soft acids and bases. J. Amer. Chem. Soc.  85: 3533-9.

Schlekat, C.E., B-G. Lee, and S.N. Luoma. 2002. Dietary metals exposure and toxicity to
aquatic organisms: Implications for ecological risk Assessment In: Coastal and Estuarine
Risk Assessment. M. Newman [Ed.]. CRC Press: Boca Raton.

Schwarzenbach, G.  1956. Organic complex forming compounds. Experientia (Suppl 5): 162-
192.

Sposito, G. 1989. The Chemistry of Soils. Sections 13.1 and 13.2. Oxford University Press,
New York, New York.

Templeton, D.M., F. Ariese, R. Cornells, L-G. Danielsson, H. Muntau, H.P. Van Leeuwen, and
R. Lobinski. 2000.  Guidelines for terms related to chemical speciation and fractionation of
elements.  Definitions, structural aspects, and methodological approaches. Pure Appl. Chem.,
72(8): 1453-1470.

Chapter U.S. EPA.  2004. Papers Addressing Scientific Issues in the Risk Assessment of
Metals. EPA/630/R-04/118 U.S. EPA Risk Assessment Forum, EPA National Center for
Environmental Assessment, Washington,  D.C-
http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=86119

U.S. EPA SAB. 2002. Review of Metals Action Plan; An EPA Science Advisory Board Report.
KPA-SAB-EC-LTR-03-001- U.S. EPA Science Advisory Board, Washington, D.C.
http://www.epa.gov/sab/pdf/ecl03001 .pdf

U.S. EPA SAB. 2000a.  Review of an Integrated Approach to Metals in Surface Waters and
Sediment. KPA-SAB-EPEC-00-005.  U.S. EPA Science Advisory Board, Washington, D.C.
http://www.epa.gov/sab/pdf/epcc0005.pdf
                                          57

-------
U.S. EPA SAB. 2000b. Review of the Biotic Ligand Model of the Acute Toxicity of Metals.
EPA-SAB-EPEC-00-006.  U.S. EPA Science Advisory Board, Washington, D.C.
http://www.epa.gov/sab/pdf/epec0006.pdf

Wallach, S. 1985. Clinical and biochemical aspects of chromium deficiency. J. Am. College
Nutr. 4:107-120.

Wang, W-X., N.S. Fisher, and S.N. Luoma. 1996. Kinetic determinations of trace element
bioaccumulation in the mussel, Mytilus edulis. Mar. Ecol. Progress Series, 140: 91-113.
                                         58

-------
Appendix A. Detailed Comments and Suggested Technical Corrections in Response to
Charge Question 3.1

Framework Section 3.1.1 - Fate and Transport.

    •   The fate and transport section (Section 3.1.1) of the human health risk assessment
       recommendations provided in the Framework currently refers to the ecological and
       environmental chemistry sections for recommendations in this area.  The SAB finds that
       this is appropriate.

Framework Section 3.1.2.1 - Background.

    •   Recommendation 1 (page 3-3, line 9) in Section 3.1.2.1 regarding background exposures
       should be modified such that the word "ambient" replaces the word "background."  In
       support of this word change, the following definition for the word "ambient" should be
       added to the glossary section:

       Ambient Levels: The amount of metals occurring in soil, water, sediment, or air that
       represent the combined contributions from natural and various anthropogenic sources.
       This ambient level may be highly region-specific but can be used as a baseline against
       which elevated levels from other natural or anthropogenic sources can be compared.

    •   Recommendation 2  (page 3-3, line 15) in Section 3.1.2.1 should be expanded by the
       addition of the following phrase: "It is also important to consider speciation of the metals
       wherever and whenever possible."

    •   The following  additional recommendation should be included in this section: "Ranges
       rather than averages should be used in risk assessments, especially for site specific
       evaluations."

 Framework Section 3.1.2.2 - Air Pathways and Inhalation Exposure.

    •   Recommendation 1 in  Section 3.1.2.2 (page 3-3, line 32) should be revised.  Particulate
       matter that is less than 2.5 micrometers in size (PM25) and nanoparticles are now of
       critical concern for the exposure and delivery of metals to humans and should be added
       as separate entities at the end of this recommendation. Support for the recommendation
       in Section 3.1.2.2 to focus inhalation exposure only on the small particles (PMio) is given
       in Atmospheric Behavior and Chemistry Section (4.1.7) where the long atmospheric
       lifetime of small compared to large particles in  the atmosphere is discussed.  In general,
       the section on atmospheric chemistry of metals is rather short and not comprehensive but
       it does support the recommendation.  EPA may want to consider addressing in this
       recommendation other larger size classes that can be important for long range transport
       and subsequent adverse effects.  However, these considerations were not addressed in
       Section 4. In order to do so, the discussion will have to be expanded.

    •   A new recommendation should be added regarding the need to consider other particle
                                           A-l

-------
       characteristics in addition to size, such as surface properties, solubility, and particle
       chemistry. The characteristics of inhaled particles are critical determinants how they
       react with biological membranes and can affect the efficacy of the uptake of metals
       across those membranes.

    •   Another recommendation should be added to the Framework regarding the need to
       consider the biological effects associated with inhaled mixtures such as metals in
       combination with other airborne pollutants including gases such as ozone (which can
       alter the permeability of the cell membrane so as to increase metal uptake by the cells).
       In addition, paniculate matter (PM) itself is a unique mixture of metals, other inorganic
       compounds such as sulfates, and organic compounds (e.g., PAHs) adsorbed onto solid
       carbon cores, and should be addressed as such.

    •   The SAB finds that recommendation 2 (page 3-4, line 1) in Section 3.1.2.2 is appropriate
       as written.

Framework Section 3.1.2.3 - Soil, Dust and Dietary Exposure Pathway.

    •   The first recommendation in this section should be deleted.  Although of less concern
       than other exposure pathways, dermal exposure should be considered because of potential
       skin effects.

    •   Recommendation 2 (page 3-4, line 16) in Section 3.1.2.3 should be revised starting at line
       20 (page 3-4) to read "consider dermal sensitization, contact dermatitis and other direct
       skin effects. For example, nickel and chromium are both common allergens in sensitized
       people (approximately 2-5% of the population for each metal), and arsenic can cause both
       local irritation as well as increased risk of cancer at sites of repeated high dose
       application.  Although dermal exposure in general is of less concern for metals, the
       potential skin effects of some metals should be considered by the risk assessor in the
       overall health evaluation."
   •   Recommendation 3 (page 3-4, line 23) in Section 3.1.2.3 is acceptable to the SAB.

   •   Recommendation 4 (page 3-4, line 27) in Section 3.1.2.3 should be modified by deleting
       text starting on line 28 (page 3-4) at the semicolon to end of paragraph (line 31). The
       SAB suggests this modification because, depending on the exposure situation, specific
       metals/metal forms, skin conditions, and dermal effects can be an issue.  Assessors
       should be aware of potential uptake of metals in specific forms (e.g., nanoparticles),
       potential uptake of metals via unique exposure conditions (e.g., bathing, showering,
       swimming), and the uptake of metals through damaged skin (e.g., irritated skin, sunburn).
       Co-exposures of metals with other toxicants can also affect dermal uptake. Dermal metal
       exposures can produce allergic dermatitis (e.g., chromium, nickel, gold), irritation (e.g.,
       arsenic, chromic acid), and skin cancer (e.g., arsenic) under certain exposure conditions.
                                           A-2

-------
Framework Section 3.1.2.4 - Water Pathway and Oral Exposure.

   •   The SAB finds that recommendation 1 (page 3-5, line 12) in Section 3.1.2.4 is acceptable
       in its current form.

   •   Recommendation 2 (page 3-5, line 17) in Section 3.1.2.4 should be amended to read:
       "It is recommended that site-specific assessments use measured metal concentrations
       within water distribution systems and at the tap."

   •   Recommendation 3 (page 3-5, line 20) in Section 3.1.2.4 should be amended by changing
       the word "negligible" to "less important." The term "surface" should be deleted.

Framework Section 3.1.2.5 - Integrated Exposure Approaches.

   •   Recommendation 1 (page 3-5, line 35) in Section 3.1.2.5 should be amended to indicate
       that the Integrated Exposure Uptake Biokinetic (IEUBK) Model should be "considered"
       rather than "recommended" and should make use of all available site-specific data, in
       particular factors that may influence oral uptake such as nutritional status of the affected
       population.

   •   With regard to recommendation 1  in Section 3.1.2.5, the SAB finds that the IEUBK
       Model is not applicable for all metals and, thus, similar models should be developed for
       other toxic metals/metalloids of concern. This is a research need that could be identified
       in a separate "research needs" document.

Framework Section 3.1.2.6 -  Bioavailability.

   •   The SAB finds that recommendations 1 , 2, and 3 (page 3-6, lines 19, 23, and 28) in
       Section 3.1.2.6 are acceptable in their current forms.

   •   The SAB finds recommendation 4 (page 3-6, line 32) in Section 3.1.2.6 to be acceptable
       in its current form for lead, arsenic and potentially  other metals.

   •   The SAB suggests deletion of recommendation 5 (page 3-6, line 35) in Section 3.1.2.6.
       The SAB notes that this is actually a research need and not a recommendation.  Research
       needs could be addressed in a separate document.

Framework Section  3.1.3.1 -  Physiologically Based Pharmacokinetic (PBPK) and
Pharamcodynamic (PBPD) Modeling.

    •   Recommendation 1 (page 3-7, line 16) should be amended by replacing "bone" with
       "storage compartments such as bone." This change de-emphasizes bone and makes a
       more general recommendation that encompasses other metals.

    •   Recommendation 2 (page 3-7, Sine 21) should be amended by deleting "and" on line 21
       and adding "(4) bioavailability, and (5) routes of exposure" at the end of the sentence
                                          A-3

-------
       (line 22, page 3-7). This change is necessary because these other important factors also
       need to be included. The SAB also recommends expansion of the discussion of PBPK
       and PBPD modeling in Section 4.2.6 to include these parameters.  References cited in
       Section 4.2.6 are appropriate, but the specific information from these citations should be
       summarized and included in the section. For example, information from the O'Flaherty
       (1998) review article on metals PBPK modeling (cited on page 4-68 of the Framework)
       should be summarized.

    •  Recommendations 3 (page 3-7, line 24) and 4 (page 3-7, line 30) in Section 3.1.3.1
       should both be deleted and the following new Recommendation 3  should be added:
       "Although there is a useful PBPK model for lead, similar models for other metals are
       lacking and need to be developed and validated."

Framework Section 3.1.3.2 - Essentiality.

       The SAB accepts the recommendation in this section, but feels that additional material is
       needed in the introductory paragraph of the section. It should be stated in this section
       that, "for some metals, there may be an apparent discrepancy between the Recommended
       Daily Allowance (RDA) and the calculated Reference Concentration (RfC) or Reference
       Dose (RfD). The EPA should consider the RDA for essential metals when considering
       the RfC/RfD. However, it should be noted that the RDA is typically satisfied by normal
       dietary intake of food and water, and therefore the RfC/RfD value may still represent a
       potential additional body burden of that metal from other dietary or extrinsic sources."
       Phrased another way, RfD/RfC values are presented as increments to RDAs.  The SAB
       also notes that there is a need for a definition of essentiality. This definition should
       address the role of the metal in an essential physiological or biochemical process.

Framework Section 3.1.3.3 - Toxicity Testing.

    The SAB suggests the following changes to this section (page 3-8):

    •   The first sentence in this section (lines 9-12) should be changed to read as follows: "At
       least five  metals are accepted as human carcinogens - arsenic, beryllium, cadmium,
       chromium (VI) and nickel."

    •   Recommendation 1 in this section (line 22 ) should be amended by adding "with
       particular attention to route of exposure, speciation and life stage." to the end of the
       sentence.

    •   Recommendation 2 (line 26) in this section should be amended by adding, "with
       particular attention to route of exposure, speciation and life stage." to the end of the
       sentence.
   •   A new recommendation should be added to this section stating that, "Animal models for
       metal toxicity need to be selected carefully with respect to species, diet, age, and sex.
       Rats, for example, sequester some metals in their red blood cells; laboratory diets
                                          A-4

-------
       frequently fail to reflect human diets; early development and senescence are periods of
       enhanced sensitivity to toxic challenges; and, sex differences in response to both
       deficiencies and excesses are universally acknowledged."

   •   The last paragraph (lines 28-31) of the section should be deleted. The statements in the
       paragraph concerning models and toxic ity testing for assessing metals are not true, nor do
       they add any value to the section.

Framework Section 3.1.3.4 - Metals Mixtures.

   •   The SAB suggests that the opening paragraph of this section mention the importance of
       metals-organic mixtures. Also, the sentence in this section containing a statement about
       selenium being protective against arsenic with reference to Section 4.3.6 for further
       discussion should be deleted. This is not a good example.  While selenium is an
       antagonist for arsenic and has been shown to inhibit arsenic's carcinogenic effects
       (particularly in animals), humans in developed nations have a sufficient amount of
       selenium in their diet and additional  selenium could be toxic. In developing nations such
       as Bangladesh, humans may have selenium deficiency and could benefit from additional
       selenium in the diet.  In addition, the SAB suggests the following changes to the
       recommendations in this section (pages 3-8 to 3-9)

       -   Recommendation 1 (page 3-9, lines 9-11) in this section should be revised to include
           the National Academy  of Sciences/National Research Council (NAS/NRC) 1988
           report on  the toxicity of mixtures as a reference (National Research Council, 1988).
           Recommendation 1 should be replaced with the following rephrasing:

           "Metal mixtures interactions and toxicity need to be clearly demonstrated by the use
           of: a) proper experimental design (National Research Council, 1988),  b) appropriate
           plotting of diagrams, and c) rigorous statistical evaluation to demonstrate synergism,
           additivity, sub-additivity, potentiation and antagonism."

           The SAB finds Recommendation 2 (page 3-9, line 13) in this section to be acceptable.

           Recommendation 3 (page 3-9, line 13) in this section should be revised to include the
           need for identifying synergy, additivity, potentiation or antagonism using appropriate
           statistical analysis.

           Recommendation 4 (page 3-9. line 22) in this section should be revised to read as
           follows: "There are established interactions that are based on molecular mimicry as a
           mechanism of action for metals.  Future research goals should determine how
           considerations of metal mimicry may affect risk assessments and metal toxicity."

 Framework Section 3.1.3.5 - Sensitive Subpopulations and Life Stages.

    •   The SAB finds that the recommendation in this section should  be revised  to read as
        follows: "Assessors should consider subpopulations with differing sensitivities that may
                                            A-5

-------
       arise as a result of differential exposure (e.g., children ingest dirt) or susceptibility (e.g.,
       elderly, immune compromised individuals, malnourishment, gender, ethnicity, genetic
       polymorphisms, etc)."

Framework Section 3.2.1 - Fate and Transport.

   •   The SAB notes that Recommendation I (page 3-11, line 7) in this section is a statement
       and not a recommendation. Recommendation 1 as currently written should therefore be
       inserted as part of the supporting text.

   •   Recommendation 2 (pe.ge 3-11, line 14) in this section is discussed in Section 4 of the
       Framework (Section 4.1.6.3.1  and Section 4.4.1.1.1). However, the treatment of this
       Recommendation in Section 4 does not provide enough detail to support the
       recommendation. The SAB finds, however, that Recommendation 2 in Section 3.2.1 is
       appropriate.

   •   The details of Recommendation 3 (page 3-11, line 19) in this section are discussed in
       Sections 4.4.1.2 and 4.4.1.1.1 of the Framework.  Recommendation 3 may be an
       important recommendation, but it is not clearly articulated from the accompanying
       support material.  In addition, the recommendation is very long (almost longer than the
       supporting text). Recommendation 3 should be shortened and supporting comments
       should be put back into the main text. The SAB also notes that a linkage in these models
       with carbon cycling is potentially important in understanding the cycling and ultimate
       effects of metals.

   •   Recommendation 4 (page 3-11, line 34) in this section addresses the use of chemical
       equilibrium models such as M1NTEQ. The utility of computer based chemical speciation
       models like MINTEQ for characterizing forms of metals is given in section (4.1.4.1.2 and
       4.1.6.4.1).  The SAB finds that adequate support is provided for this recommendation.

   •   Recommendation 5 (page 3-11, line 38) in this section is discussed in Section 4 of the
       Framework, starting on page 4-99, and is consistent with EPA policy. The SAB notes,
       however, that Recommendation 5 is not written as a recommendation, but rather as a
       statement.  Recommendation 5 should be shortened and re-stated in the form of a
       recommendation.
      The SAB finds that Recommendation 6 (page 3-12, line 6) in Section 3.2.1 of the
      Framework is well supported by text provided in the environmental chemistry section of
      the document.  This recommendation is supported in the discussion of the limitations of
      solution speciation computer based models (Section 4.1.6.4.2). This issue is also given
      some support in the discussion on the limitation of the equilibrium partition approach in
      the discussion of limitations (Section 4.4.1.1.2) of the aquatic transport models. This
      limitation is certainly one of the most important for modeling.

      Recommendation 7 (page 3-12, line 12) of this section is supported, but not in the section
      of the Framework that is referenced.  Rather, the limitation of the equilibrium partition
                                         A-6

-------
       coefficient and equilibrium approach is given in the discussion of limitations of the
       aquatic transport models. This discussion is provided in Section 4.4.1.1.2 but not in
       Section 4.1.4. However, the equilibrium assumption for modeling metal partitioning to
       and from aged soils is a limitation that is not mentioned in Recommendation 7. The SAB
       notes that metals are not likely to be in readily reversible associations with solid phases in
       aged soils. This point on aging is made in Section 4.1.6.3. Recommendation 7 should
       not stand alone as a recommendation, but rather be included in the supporting text. The
       SAB notes that Recommendation 6 in Section 3.2.1  can encompass the issue of partition
       coefficients.  The SAB also notes that there should be less emphasis given to static
       "partition coefficients" and more emphasis on dynamic partitioning processes.

   •   Support for Recommendation 8 (page 3-12, line 22) of this section is provided throughout
       the document in the discussion of the importance of oxidation state changes for certain
       metalloids, and in the environmental chemistry section (Section 4.1).  For example, the
       environmental chemistry section on the importance  of pH and redox conditions (Section
       4.1.3.2), and the atmospheric behavior/chemistry section (Section 4.1.7) address
       Recommendation 8. However, it may be inappropriate to change input parameters to
       overcome the shortcomings of the process modeled.  As such, the SAB suggests
       removing this recommendation and the text on page 4-101 (lines 25-33) from which it
       was taken.

Framework Section 3.2.2 - Water Column Exposure, Bioavailability and Effects

   •   The SAB finds that the recommendations contained in this section were generally well-
       stated and well-supported in Section 4.

Framework Section 3.2.3 - Background.

   •   The SAB finds that the recommendations in Section 3.2.3 (background) are generally
       supported by the text in Sections  4.5.4.1-4.5.4.2  of the Framework, but statements in
       Section 3.2.3 are not really "justified" by data (i.e., graphs showing variability). This
       type of "support" is probably not absolutely necessary, but in a strict sense, the
       Framework fails to justify the conclusion in Section 3.2.3 with data.

   •   The recommendation addressing the importance of considering background
       concentrations in metals risk assessment is discussed in various places throughout the
       document. It is identified as a key metal issue in the problem formulation and metals
       principles section (Section 2) and given specific emphasis in the human exposure
       pathway analysis section (Section 4.2.2.1). Background concentration effects are also
       discussed in the section on characterization of ecological effects (Section 4.5.4).  The
       SAB notes that in all the recommendation statements provided in the Framework, it
       would be useful to indicate where the information is discussed in Section 4.

   •   Background metal concentrations issues are not discussed in the environmental chemistry
       section (Section 4.1) of the Framework. The SAB notes that in the environmental
       chemistry part of the document it would be useful to include a section on the natural
                                           A-7

-------
       occurrence of metals.  It would be useful to highlight in the environmental chemistry
       section those metals and regions for which background concentrations would be
       important.  The issue of aging as discussed in Section 4.1.6.3 is suggestive of the relative
       importance of background versus recent metal inputs into soils and sediments and its
       implied significance to bioavailability and mobility.  Similarly, a discussion of the effect
       of early diagenetic reactions on the fate and effects of metals would be helpful.

    •  The SAB finds that the relationship between the recommendations in Section 3.2.4 on
       bioaccumulation and the support in Section 4.5.8 is muddled by the lack of a clear
       presentation and consistent use of definition of "bioaccumulation factor" and
       "bioconcentration factor" (BAF/BCF).  Once BAF/BCF are clearly defined and used
       consistently, it will be possible to assess these sections critically.

Framework Section 3.2.4 - Bioaccumulation.

       The SAB finds that the recommendations in this section are unclear,  contradictory,
       inconsistent, and ill supported. As discussed in the responses to charge questions 3.11,
       3.12, and 3.13 below, Section 4 of the Framework does not adequately reflect the
       recommendations in Section 3.2.4 concerning bioaccumulation. In general, the SAB
       feels the EPA needs to revise the recommendations in this section to increase clarity and
       conciseness. For example, EPA should consider: 1) combining and reconciling
       Recommendations  1 and 3 (page 3-17, lines 16 and 27) in this section; 2) Combining and
       clarifying Recommendations 4, 5, (page 3-17 lines 31 and) and 8 (page 3-18, line 12) in
       this section; and 3) Combining Recommendations 6 and 7 (page 3-18, lines 1 and 5) in
       this section. The issue of diet must be reflected as a route of exposure in the revision.
       The SAB finds that Recommendation 9 (page 3-18, line 16) in this section can stand as
       drafted.

Framework Section 3.2.5 -Trophic Transfer, Biomagnification, and Dietary  Toxicity.

    •   The SAB finds that Recommendation 1 (page 3-19, line 9) in this section of the
       Framework needs to reflect the importance of trophic transfer. It is suggested that the
       statement be revised by adding the phrase: "...classifying hazards or risks of inorganic
       metal compounds, [whereas, trophic transfer should be considered]."  The SAB finds that
       Recommendation 2 (page 3-19, line 17) in this section should be revised to be more
       directed and concise.  As written, the recommendation is contained in the 1st sentence.
       The remainder of the text is clarifying information and should be moved into the
       supporting text description.

Framework Section 3.2.6 - Sediment Exposure and Effects.

    •   The SAB finds that all of the recommendations in this section of the Framework should
       be reconsidered  in light of the discussion contained in the response to charge question
       3.14 below.
                                          A-8

-------
Framework Section 3.2.7 - Metals Mixtures,

   •   The SAB finds that the discussion in Chapter 4 that is related to the recommendations in
       Section 3.2.7 of the Framework should be further developed.  EPA should consider the
       addition of a recommendation to address the inclusion of empirical studies of metals
       mixtures in the field. The SAB finds that the "field" part is left out of the current version
       of the Framework.  Finally, the SAB finds that the concept of Quantitative Ion Activity
       Relationships (QICAR) is not well developed in Sections 3 or 4. EPA needs to justify
       that this concept is sufficiently well developed and validated to be included in such a
       specific recommendation.

Framework Section 3.3.1.1 - Atmospheric Chemistry and Behavior.

   •   The SAB finds that Recommendation 1  (page 3-22, line 18) in this section is not a
       recommendation. It should therefore be removed and included as commentary in the
       paragraph description.  The SAB also notes that the Community Multi-Scale Air Quality
       (CMAC) Model is not mentioned by name in Section 4.1.7 as implied by the referencing
       to Section 4.1.7.

Framework Section 3.3.1.2 - Soil Mobility.

   •   The SAB finds that Recommendation 1  (page 3-23, line 6) in this section is simply a list
       of measurement techniques and not necessarily a recommendation. Little supporting
       information is given on the utility of each  technique and how the information may be
       useful in a risk analysis context.  Some chemical techniques and speciation tools are
       covered in the referenced environmental chemistry section (Section 4.1.8).  However, if
       the point to be made in this section of the Framework is that these tools should be used to
       help with site specific assessment of metals, and to provide guidance on relative mobility,
       then this should be stated in Recommendation 1. The SAB also notes that little
       information is provided in the Framework about analytical chemical methods that are
       currently commonly used for metal ion  speciation.

   •   The SAB finds that Recommendation 2 (page 3-23, line 15) in this section is not a
       recommendation but a statement.  Supporting information on the need to use computer
       models for predicting speciation changes in soil solutions is provided in the
       environmental chemistry section  (Section  4.1.6.4.1.1.). The SAB suggests that an
       appropriate statement to he included in Recommendation 2 would be that computer
       speciation models should be considered when: 1) a more definitive analysis of the impact
       of metal ion speciation in metal risk assessment is required for site specific  level risk
       characterization, 2) the model assumptions are appropriate for the application, and 3)
       sufficient site characterization data are available.

   •   The SAB finds that Recommendation 3 (page 3-23, line 20) in this section is well
       supported in the Framework. However the recommendation is not stated in the form of a
       recommendation.  The SAB suggests that Recommendation 3 might be restated to
       emphasize that Kd values can only be used when they are either calibrated for a specific
                                          A-9

-------
 site, or have sufficient functionality built in to account for the variability of K
-------
      provided in the environmental chemistry section. As stated previously, one must also
      account for changing environmental conditions.

   •  The SAB finds that Recommendation 7 (page 3-24, line 5) in this section is not a
      recommendation. Discussion of the Generalized Two-Layer Model (GTLM) is provided
      in Section 4.1.4.1.2 of the Framework.  However, support for the requirement of isotropic
      fluid flow and fast, reversible, and linear sorption is not given in the environmental
      chemistry section of the Framework. The inherent assumption of isotropic fluid flow,
      however, is common to transport models. While the need for fast and reversible sorption
      is true, linear sorption is not required per se.  One of the attributes of the GTLM is that it
      can account for the nonlinearities in sorption as a function of pH and changing amounts
      of solid to liquid. The SAB notes that little discussion was provided in the Framework on
      the conditions under which the use of models is appropriate. The SAB suggests that a
      more elaborate discussion of the limitations (data or field conditions) be added to the
      metal sorption section to describe the types of scenarios where such models are
      appropriate. In general, the SAB finds that the discussion in the Framework of sub-
      surface transport is limited.

   •  Recommendation 7 (page 3-24, line 10) in this section of the Framework addresses the
      PHREEQC model. It is not clear why a separate recommendation is needed concerning
      PHREEQC unless the point to be made is that three dimensional models are also now
      available that couple metal ion surface complexation models with transport.  The SAB
      finds that Recommendation 7 is largely a repeat of the same sentence from Section
      4.1.4.1.2 of the Framework.  The sentence is given there without further discussion.

Framework Section 3.3.2.2 - Soil Invertebrates and Plants.

   •  The SAB finds that the recommendations in this section of the Framework should be
      revisited and revised in light of comments in the response to charge question 3.11 below.

Framework Section 3.3.2.2.1 - Soil Invertebrates.

   •  The SAB finds that the recommendations in this section of the Framework are well-stated
      and well-supported.

Framework Section 3.3.2.2.2 - Plants.

   •  The SAB finds that the recommendations in this section of the Framework (as drafted)
      should be reduced in scope such that the actual recommendations are clearly stated and
      the explanatory statements are moved to the supporting text. The text in
      Recommendation 1 (page 3-28, line 33) in this section, reflecting the soil plant barrier
      concept, needs to be shortened but expanded in supporting paragraphs. Recommendation
      3 (page 3-29, line 12) in this section, discussing the issue of aerial deposition, should be
      reconsidered and dropped or revised to reflect supporting information. The SAB finds
      that this recommendation is not adequately supported by text. The SAB finds that
      Recommendation 4 (page 3-29, line 16) in this section  is a statement not a
                                         A-il

-------
       recommendation and should be moved to the supporting text.

Framework Section 3.3.2.3 - Wildlife.

    •  The SAB finds that the recommendations in this section of the Framework are well
       defined and adequately supported. It is suggested that Recommendation 5 in this section
       be revised as follows: 'Although bioaccumulation and trophic transfer of metals does
       occur [and should be considered], with some exceptions (e.g., selenium and mercury)
       biomagnification (i.e., increases in concentration through the food web) is a less
       important consideration and may be assumed to be unimportant." Recommendations 3,
       4, and 5 in this section should be combined into a single recommendation.
       Recommendation 5 in this section contains a reference to the general scientific literature.
       This should be relocated to another part of the document.

Framework Section 3.3.2.4 -Food Chain Modeling.

    •  The SAB finds that the recommendations in this section should be revised to make them
       more concise. Recommendation 2 (page 3-31, line 9) in this section of the Framework is
       not a recommendation and should be moved to the supporting text. Recommendations 3,
       4, and 5 (page 3-31, lines 11, 25, and 29) of this section should be consolidated into a
       single recommendation.

Framework Section 3.3.2.5 - Bioaccumulation.

    •  EPA should reconsider and re-evaluate the recommendations in this section in the light of
       previous comments, and make sure that parallels between  soils and sediments are
       developed.

Framework Section 3.3.3.1 - Adaptation and Acclimation.

    •  The SAB finds that there is confusion about what is intended in the Framework by the
       term "acclimation."  It is unclear whether EPA is addressing the question of "true"
       metals acclimation and the resulting increase in tolerance and/or resistance, or suggesting
       that care should be taken in culturing organisms for testing to ensure that they are not
       "overly sensitive" owing to the fact that they were raised in metals-deficient conditions.

Framework Section 3.3.3.2 — Essentiality.

    •  The SAB finds that Recommendations  1 (page 3-35, line 23) and 5 (page 3-36, line 2) in
       this section be removed  and incorporated into the supporting text of the document
       because they are not recommendations, but rather informational statements.

Framework Section 3.3.3.3 - Metals Mixtures.

    •   In general, the SAB finds that the metals mixtures recommendations in this section of the
       Framework are adequate. However, the SAB notes that there is a need to be mindful of
                                         A-12

-------
       the importance of evaluations conducted in the "real world."

Framework Section 3.3.3.4 - Toxicity Testing.

   •   The SAB finds that the recommendations for toxicity testing and extrapolation of effects,
       as developed for terrestrial ecosystems, need to be developed and included in the aquatic
       section of the Framework. Toxicity testing of metals has strengths and limits as an
       assessment tool metals.  For example, limits derive from: the use of surrogate species
       versus the diversity of responses to metals, among metals and among species; and the
       lack of dietary exposures in the toxicity testing data bases usually used by risk assessors.
       There are unique effects of metals that are well known in some aquatic environments
       (e.g., stream insect communities; selenium and mercury effects on upper trophic levels)
       and poorly known in others. In light of these considerations, the SAB finds that the
       recommendations in this section are not well articulated with regard to evaluation of
       national and site specific risk from metals.  The recommendations contained in this
       section need to be concise and explanatory text needs to be moved into the supporting
       body of text. The SAB finds that the recommendations in the section were generally
       supported by the text in Section 4  of the Framework.

Framework Section  3.3.3.5 - Extrapolation of Effects.

   •   The SAB finds that actual recommendations need to be made and "statements" moved
       into the text. For example,  Recommendation 4 in this section is a statement. The SAB
       finds that the recommendations  in this section are not well-supported by information in
       Section 4 of the  Framework.

Additional, Specific Comments on Section 3 of the Framework

 The SAB provides the following additional specific comments on the metals risk assessment
recommendations section of the Framework document. The pertinent pages and line numbers in
Section 3 of the Framework are referenced below.

   •   Some of the recommendations in Section 3 of the Framework are not sufficiently specific
       to be useful. On page 3-! I, for example, the following recommendation is made about
       use of chemical  equilibrium models: "Most of the available transport models do not
       currently include chemical speciation subroutines.  In such cases, chemical equilibrium
       models such as M1"NTEQ serve as useful alternatives for characterizing the forms of
       metal that are present."  This statement is not incorrect, but it is not clear how chemical
       equilibrium models can be used to consider speciation in transport assessments.

   •   Page 3-10, lines 31-32: In light of discussions in Section 4 of the Framework and in the
       issue paper on the environmental chemistry of metals, it would be more appropriate to
       state that partitioning (and not partition coefficients) are important. This statement
       should be followed by a discussion of how chemical speciation calculations are preferred
       in determining metal partitioning, but in situations  where sufficient data and modeling
       tools arc not available, partition coefficients should be assigned with great care to account
                                          A-13

-------
    for the effects of pH, inorganic and organic ligand concentrations, competitive
    interactions, and redox chemistry. Although the comments on partitioning and partition
    coefficients may seem minor, it is important that the EPA begin to move away from the
    paradigm of partition coefficients for metals and place greater emphasis on the more
    appropriate concept of metal speciation.  The attention of risk assessors should be drawn
    more directly to the use of the MTNTEQ model to relate metal speciation to soil-water or
    particle-water partition coefficients, such as one would use in a transport model. In
    addition, risk assessors must be advised to carefully investigate the effects of changes in
    key environmental parameters, such as pH, DOC, and particle composition on metal
    speciation and partition coefficients, whether they are calculated using MINTEQ or
    obtained by measurement. In this way, a commonly available tool, MINTEQ, can serve
    as both a conceptual and practical bridge between the more rigorous concepts and
    approaches implied by "metal speciation" and the more familiar concepts and empiricism
    of the "partition coeffic ient."

    Page 3-11, line 24: In complex models, organic carbon cycling should specifically be
    included to account for temporal  and spatial changes in particulate organic carbon (POC),
    dissolved organic carbon (DOC), redox conditions, and for Hg assessments, sulfate
    reduction rates.  It is therefore recommended that EPA add organic carbon modeling to
    line 24 (e.g., as "hydrodynamic, sediment transport, organic carbon cycling, and chemical
    transport algorithms").

    Page 3-12, line 12-20: The focus  should be on metal partitioning (and not partition
    coefficients).  In addition to the recommendation for further consideration of equilibrium
    assumption, it may be even more important to recommend an appropriate approach for
    calculating metal partitioning from chemical speciation calculations, and when sufficient
    data and modeling expertise is not available, to state what factors need to be considered
    in assigning a partition coefficient.

    Page 3-14, lines 36-37: Quantitative Ion Character Activity Relationships (QICARs)
    appear to be an important tool for extrapolation of metal availability and toxicity data.
    However, the detailed discussion  of QICARs on pages 4-153 and 4-154 is very brief and
    could be further developed.

    Page 3-18, lines 1-3: Discussions in Section 4 of the Framework  on biotic ligand models
    focus on bioavailability and toxicity from metal binding at the gill. The SAB notes that
    there are no discussions on how biotic ligand models have been used in estimating
    bioaccumulation.
•  Page 3-19, lines 9-15: In discussing the rarity of metal bioaccumulation, a qualifying
   statement should be added for the methylmercury and organoselenium exceptions.

•  Page 3-21, lines 22-23: It is not clear from discussions in Section 4 of the Framework that
   the BLM has been applied to metal mixtures.

•  The recommendations in Section 3 of the Framework are often given without the
                                      A-14

-------
precautionary statements that were part of discussions in Section 4 of the document. For
example, Section 4 indicates the limitations of several approaches when applied to clay-
rich sediments and soils.

In general, all of the "recommendations" under soil mobility (Section 3.3.1.2) need to be
reworded and stated in the form of recommendations or it should be stated at the
beginning of Section 3 of the Framework that the lists are guidance statements or
recommendations rather than just recommendations.

Although organo-metal transformation processes are discussed in the environmental
chemistry section of the Framework (Section 4.1.9), the recommendations at the end of
Section 3.3.1.3 (transformation in soils, page 3-25) are not taken directly from the
information provided. Any recommendation listed in Section 3, should follow naturally
from the information and context provided in Section 4.  The summary paragraph in
Section 3 and the  recommendations listed in Section 3.3.1.3 do not seem to be taken from
the Section 4 summary on organo-metal transformations.

Page 3-10, lines 23-25: Regarding model complexity, the statement that more complex
models are not necessarily better gives no basis for decision. Calibration is arguably the
key issue in making this choice and discussion of this point would be helpful.

Page 3-11: The discussion of partitioning seems out of place given critique of partitioning
that is provided later in the Framework.

Page 3-14: Hardness (competing cationic metals) is a factor that is independent  from
speciation. The suggestion to only use it when speciation data are not available does not
make sense.

Section 3.2.3:  Default use of state averages for backgrounds would be erroneous if non-
point sources are  significant in comparison to geological sources. The SAB notes that
this could be ascertained on a metal-by-metal basis prior to adopting state average as a
background.

Section 3.2.4:  The discussion in the Framework concerning the appropriate use  of
BCF/BAF is confusing. Paragraphs at the bottom of page 3-16 and top of page  3-17
seem to offer conflicting statements concerning the use of BCF/BAF.

Page  3-17, lines 8-9: The statement concerning whole body concentration and potential
for toxicological impact is likely to be true, but the question of correlation between whole
body concentrations and concentrations in specific organs/sites should be considered.

Section 3.2.6: Recommendation of the use of the SEM-AVS approach without
considering other approaches is neither balanced nor justified
                                    A-15

-------
 Appendix B. Speciation

   Among risk assessors and scientists working on the environmental chemistry and
 ecotoxicology of metals, the concept of "chemical speciation" Is fundamental. Despite this fact,
 or perhaps because of it, a variety of context-specific uses of the term, along with the related
 term "chemical species," have developed. This practice can confuse newcomers to the field,
 perhaps even hindering their apprehension of concepts that are not in themselves difficult. To
 remedy this situation, the SAB recommends that the environmental chemistry section begin with
 a set of definitions adapted from recent IUPAC recommendations (Templeton et a!., 2000).
 Quotations from this source are in italics.

 Species: Chemical compounds that differ in isotopic composition, conformation, oxidation or
 electronic state, or in the nature of their complexed or covalently bound substituents, can be
 regarded as distinct chemical species.

 In environmental chemistry, the phase the species occurs in - gas, liquid, aqueous solution,
 mineral, or adsorbed on  an interface between phases - generally is also specified in a complete
 definition.

 Note that this definition  applies equally to the environmental chemistry of organic compounds
 and of metal ions, although there are important differences in how the term is used in practice.
 In the context of the environmental chemistry of metals, chemists speak of a metal species as a
 "specific form of an element defined as to isotopic composition, electronic or oxidation state,
 complex or molecular structure " and phase. In the context of environmental organic chemistry,
 chemists do not usually refer to an organic compound as specific form of carbon, although every
 organic compound is one. Rather, as long as its core structure remains intact, each different
 protonation state, complex of a metal ion,  and occurrence in different phases of an organic
 compound may be  referred to as a different species of the compound.

 Speciation: According to the above definition  of species, it is apparent that the reactants and
 products of any properly written chemical reaction are distinct chemical species. Indeed, the
 concepts of species and reactions are intimately related since any process that brings about a
 chemical change by definition results in the formation of a new species.  As a result of this
 logical relationship, and  possibly also its parallel to the concept of evolutionary "speciation" in
 biology, some geochemists and environmental  chemists have "applied the word speciation to
 describe the transformations taking place during cycling of the elements. " However, the IUPAC
 has recommended against this  use of "speciation," instead suggesting the term species
 transformation. Given its consistency with the usage of "transformation" in the field of
environmental organic chemistry, this recommendation should be easily accepted and put into
practice.

The IUPAC also recommends  against using the term speciation to indicate the analytical activity
of identifying chemical species and measuring their distribution.  Sometimes, it is used to
indicate that a method gives more information on the form in which the element is present than
other more commonly applied techniques (e.g., measuring distinct organomercury compounds as
opposed to a total mercury determination). In order to avoid confusion, [IUPAC] recommends
                                          B-l

-------
using the term speciation analysis when referring to the analytical activity of identifying and
measuring species.

Instead, the lUPAC-recommended use of speciation is the distribution of an element amongst
defined chemical species in a system. Normally, a quantitative description of the speciation of an
element is implied.  Such a distribution could be the result of: i) one or more chemical analyses
of a sample, ii) chemical modeling of a laboratory solution of known composition, or iii)
chemical modeling of an environmental system. When not clear from the context, the terms
analytical speciation and modeled speciation may be helpful in distinguishing these methods
used to obtain the speciation.

As a practical matter, the degree of resolution adopted in any description of the speciation of a
system will depend on:

i)      the relevance of the species differences for our understanding of the system under study,
ii)     our ability to distinguish between the various species analytically,
iii)     our ability to model the speciation in some operationally-defined or experimentally-
       controlled fraction of an analyzed substance.

While  some analytical methods directly determine the concentration of a single species in an
environmental sample or matrix, most common environmental analyses measure several related
species, or fractions. IUPAC recommends that the process of classification of an analyle or a
group of analytes from a certain sample according to physical (e.g., size, solubility) or chemical
(e.g., bonding, reactivity) properties undertaken by a chemical analyst be referred to as
fractionation.
                                           B-2

-------
Appendix C. Suggested Editorial/Wording Changes

Section 2 of the Framework

    •  Section 2.1.7 discusses toxicity testing and implies that toxicity is the metal impact of
       primary concern. However, the SAB notes that metal effects on the environment can be
       much broader than effects measured in a toxicity test endpoint (e.g., long-term impacts
       on ecosystem structure). The SAB therefore suggests that the factor be re-named and
       discussed as "toxicity." The terrestrial part of Section 3 (Sections 3.3.3.4 and 3.3.3.5)
       extends "toxicity testing" to include "extrapolation to effects" (in nature). The SAB
       recommends that the problem definition of "toxicity" in Section 2 be clarified in a
       similar way. It is important to take into account limits and linkages between  toxicity
       testing and adverse effects.  Both Section 4 and EPA's Metals Issue Papers include
       useful discussions of effects of metals on populations and communities of organisms.

    •  The linkage of Figure 2-3 to the text could be enhanced by modifying the footnote box
       "Key Metal Issues" in Figure 2-3 to include references to specific subsections in the text.
       The footnote box should be reconsidered to determine how well it clarifies the figure and
       relates the figure's components to the text. The SAB suggests that the footnotes to
       Figure 2-3 might be improved by listing just the key factors that impact the conceptual
       model components shown. The SAB offers the following specific comments on Figure
       2-3.

          The footnotes to Figure 2-3 would be easier to understand if the words were not
          abbreviated in the description of Ml through M9 in the figure legend.

          The footnote referring to Ml of Figure 2-3 should include organic carbon  cycling.

          The meaning of "concentration dependency" in the footnote referring to M2 of Figure
          2-3 is  unclear.

          In the blocks on Figure 2-3, the word "chemical" should be changed to "metal".

   •   The conceptual model represented in Figure 2-3 was developed to describe the
       assessment of classes of metals identified in Table 2-1 in Section 2 in the Framework.
       The SAB  offers the following specific comments on the lists of metals in Table 2-1:

          Mg is  an essential metal and should be added to Table 2-1.

          Silicon is in Table 2-4 but not in Table 1-2 of Section  1. For consistency,  these  tables
          should have the same elements.

          It is unclear why the particular metals in Tables 1-2 and 2-1 were selected  to be
          included in the tables, and why others were omitted. Some comment should be
          included concerning risk assessment for other metals such as tungsten, uranium, or
          tellurium that  may be important in local, regional, or national settings. This is
                                          C-!
J

-------
          discussed in lines 9 to 13 of page 1-3 in the Framework, but the relevance to all
          metals should be repeated in introducing Table 2-1.

       -   The SAB suggests that the Framework contain references to the work of authoritative
          scientific panels charged with making recommendations regarding essential metals,
          such as the National Academy of Sciences (2000). If changes occur in this field over
          time, readers can be directed to these more up-to-date sources of information. The
          SAB also notes that the following reviews by individual experts on chromium
          essentiality should be cited: Hathcock (1996),  Lukaski (1999), Mertz (1993), Mertz
          (1995), and Wallach (1985). Additional comments on the list of metals  included in
          Table 2-1, and the classifications presented there, are provided in the response to
          Charge Question 3.9.

Sections 3 and 4 of the Framework

   •   The concept of soils and the terminology associated with soil substrates needs to be
       clearly defined and the text should be revised to accommodate modern nomenclature in
       soil science for organic and mineral soil horizons and soil types.  Concepts used should
       accommodate soil substrates in urban, wetland, forested, agronomic and disturbed
       ecosystem contexts consistent with U.S. Department of Agriculture (USDA) Natural
       Resource Conservation Service (NRCS) terminology. Several soil properties (e.g., pH,
       cation exchange capacity [CEC]) are often discussed in this Section 3.3 of the Framework
       because of their appropriate importance in metal risk assessment activities.  These
       properties can be  highly operationally defined based on the methods chosen but there is
       only a passing mention in the Framework of the importance of methods. This subject
       should be explicitly developed.

   •   A statement should be included about the potential for longer-scale transport of metals
       from a source through the atmosphere to soil, water, or air, from  which exposure
       ultimately occurs. Even if the process for metals follows principles already  established
       and described for organic compounds, and EPA does not want to repeat that description
       in the Framework, a statement about the similarities and differences between inorganic
       and organic compounds would improve the completeness of the Framework. For
       example, while many metals are transported in the atmosphere primarily only on the
       surfaces of particles, many organic compounds are transported in the atmosphere
       primarily as a component of the vapor phase.

   •   A statement about the potential importance of atmospheric transport to "background"
       concentrations of metals in the environment should be included either  in the section of
       the Framework discussing atmospheric chemistry or in the "background" section.

   •   The SAB notes that, EPA has used the term, "duff", in (he terrestrial section of the
       Framework (page 3-27) when discussing factors influencing metal availability and
       accumulation.  This term is many decades out of date. The SAB therefore suggests that
       HPA delete the term and instead use the "0 horizon or litter layer." Use of the correct
       terminology is important in order to address concerns about soil  measurements and soil
                                          C-2

-------
       concepts. In a forest region, the forest floor horizon is the 0 horizon. Standardizing soil
       to the top 10 - 12 cm is not appropriate across a range of ecosystem conditions to include
       urban, wetland, undisturbed forest, agronomic, and disturbed systems.  A more
       appropriate nomenclature for soil horizons and types consistent with USDA NRCS
       terminology should be defined and used.

 The SAB notes that the following statements in Section 4 should be checked for accuracy and, as
 warranted, corrected:

    •  Page 4.3, line 1: The order of the metal sulfides appears to be incorrect. Iron should be
       moved between zinc and manganese. The solubility constants cited should be checked
       against established compilations of thermodynamic data.

    •  Page 4-40: The equation showing dimethylmercury photolysis should be checked.
       Dimethylmercury does not absorb sunlight directly and direct photolysis is unlikely.
       Formation of two methyl radicals is also unlikely. Atmospheric oxidants, however,
       would be expected to oxidize dimethyimercury (as discussed in the last paragraph on the
       page). Also, demethy! ation is unlikely to occur via sorption to particulate matter, as
       suggested in line 31 of the same page.

    •  Page 4-42, lines 15-16: The statement that 15-30 percent of arsenic is volatilized is
       almost certainly due to arsine (AsHs), rather than methylation. Thus, the statement is out
       of place in a methylation paragraph.

    •  Page 4-40: The atmospheric transformation sections appear to be written in a manner that
       is inconsistent with other sections (e.g., paragraph length and formatting).  The sections
       are very short, relative to the previous or following sections.

    •  Page 4-39, line 7: The following text should be reworded "... formation of less
       bioavailably charged metal-sulfur complexes". The SAB questions whether the metal
       sulfur complexes are actually charged.

    »  Page 4-39, line 11: Use of the word "unbioavailable" appears to  be a bit awkward.

The following specific revisions are also suggested by the SAB to improve the clarity of Section
4.2.1.

    •  Page 4-2, lines 8-10:  The introductory paragraph contains broad generalities that are not
       all strictly accurate; it should be completely rewritten.  EPA should define "acids" and
       "bases" and then state which metal species are usually acids and which ligand atoms are
       usually bases.

    •  Page 4-2, lines 12-13 and  15: Revise the document to qualify and/or provide references
       for the statement about "toxic reaction" being directly related to  the nature of the metal at
       the surface of the organism, and the statement about toxicity. This is addressed in the
       comment above on toxicity.
                                           C-3

-------
•  Page 4-2, line 16: Change "introduced" to "described" because the concept was
   introduced earlier by Pearson (1963) and others (Arhland et al., 1958; Schwarzenbach,
   1956)inthel950's.

•  Page 4-2, line 17: Delete "in this concept" because the statement is true in general.

•  Page 4-2, line 21: Delete "mobile and easily moved" to avoid confusion with oxidation-
   reduction reactions; "deformable" and "polarizable" are the appropriate terms.

•  Page 4-2: If a box is necessary to define "ligand," a box should also be used to define
   "complex."

•  Page 4-3, line 1: Delete the clause "which are less toxic;" such statements about relative
   toxicity can be included in the text, where appropriate justification and qualification can
   be given, rather than in the title where it appears without justification and qualification.

•  Page 4-3, line 2: The U.S. spelling of sulfur should be used.

•  Page 4-3, lines 2-3: The appropriate term to be used is "extent of binding," not "strength
   of binding" because the strength is intrinsic to the metal and ligand, and the pH effect is
   more accurately described as a competition effect.

•  Page 4-3, line 5: Change "many of the hard metals" to "some of the hard acids;" to avoid
   confusion, use the terminology "hard and soft acid" consistently, don't switch to "hard
   and soft metals."
                                        C-4

-------

-------